Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

Category articles

Dermatology Dermatology

Prediction of Chemosensitivity in Multiple Primary Cancer Patients Using Machine Learning.

In Anticancer research

BACKGROUND/AIM : Many cancer patients face multiple primary cancers. It is challenging to find an anticancer therapy that covers both cancer types in such patients. In personalized medicine, drug response is predicted using genomic information, which makes it possible to choose the most effective therapy for these cancer patients. The aim of this study was to identify chemosensitive gene sets and compare the predictive accuracy of response of cancer cell lines to drug treatment, based on both the genomic features of cell lines and cancer types.

MATERIALS AND METHODS : In this study, we identified a gene set that is sensitive to a specific therapeutic drug, and compared the performance of several predictive models using the identified genes and cancer types through machine learning (ML). To this end, publicly available gene expression datasets and drug sensitivity datasets of gastric and pancreatic cancers were used. Five ML algorithms, including linear discriminant analysis, classification and regression tree, k-nearest neighbors, support vector machine and random forest, were implemented.

RESULTS : The predictive accuracy of the cancer type models were 0.729 to 0.763 on the training dataset and 0.731 to 0.765 on the testing dataset. The predictive accuracy of the genomic prediction models was 0.818 to 1.0 on the training dataset and 0.759 to 0.896 on the testing dataset.

CONCLUSION : Performance of the specific gene models was much better than those of the cancer type models using the ML methods. Therofore, the most effective therapeutic drug can be chosen based on the expression of specific genes in patients with multiple primary cancers, regardless of cancer types.

Zhang Xianglan, Jang M I, Zheng Zhenlong, Gao Aihua, Lin Zhenhua, Kim Ki-Yeol

2021-May

Multiple primary cancers, cancer type, chemosensitivity prediction, gene expression, machine learning

General General

Machine Learning Pipeline for Segmentation and Defect Identification from High-Resolution Transmission Electron Microscopy Data.

In Microscopy and microanalysis : the official journal of Microscopy Society of America, Microbeam Analysis Society, Microscopical Society of Canada

In the field of transmission electron microscopy, data interpretation often lags behind acquisition methods, as image processing methods often have to be manually tailored to individual datasets. Machine learning offers a promising approach for fast, accurate analysis of electron microscopy data. Here, we demonstrate a flexible two-step pipeline for the analysis of high-resolution transmission electron microscopy data, which uses a U-Net for segmentation followed by a random forest for the detection of stacking faults. Our trained U-Net is able to segment nanoparticle regions from the amorphous background with a Dice coefficient of 0.8 and significantly outperforms traditional image segmentation methods. Using these segmented regions, we are then able to classify whether nanoparticles contain a visible stacking fault with 86% accuracy. We provide this adaptable pipeline as an open-source tool for the community. The combined output of the segmentation network and classifier offer a way to determine statistical distributions of features of interest, such as size, shape, and defect presence, enabling the detection of correlations between these features.

Groschner Catherine K, Choi Christina, Scott Mary C

2021-May-06

HRTEM, automated analysis, deep learning, segmentation, structure classification

General General

Persistent dependent behaviour is accompanied by dynamic switching between the ventral and dorsal striatal connections in internet gaming disorder.

In Addiction biology ; h5-index 43.0

Cross-sectional studies have suggested that functional heterogeneity within the striatum in individuals with addictive behaviours may involve the transition from ventral to dorsal partitions; however, due to limitations of the cross-sectional design, whether the contribution of this transition to addiction was confused by individual differences remains unclear, especially for internet gaming disorder (IGD). Longitudinal functional magnetic resonance imaging (fMRI) data from 22 IGD subjects and 18 healthy controls were collected at baseline and more than 6 months later. We examined the connectivity features of subregions within the striatum between these two scans. Based on the results, we further performed dynamic causal modelling to explore the directional effect between regions and used these key features for data classification in machine learning to test the replicability of the results. Compared with controls, IGD subjects exhibited decreased functional connectivity between the left dorsal striatum (putamen) and the left insula, whereas connectivity between the right ventral striatum (nucleus accumbens [Nacc]) and the left insula was relatively stable over time. An inhibitory effective connectivity from the left putamen to the right Nacc was found in IGD subjects during the follow-up scan. Using the above features, the classification accuracy of the training model developed with the follow-up was better than that of the model based on the initial scan. Persistent IGD status was accompanied by a switch in the locus of control within the striatum, which provided new insights into association between IGD and drug addiction.

Wang Min, Zheng Hui, Zhou Weiran, Jiang Qing, Dong Guang-Heng

2021-May-06

dynamic causal modelling, functional connectivity, internet gaming disorder, longitudinal study, machine learning, striatum

General General

EEG-based auditory attention decoding using speech-level-based segmented computational models.

In Journal of neural engineering ; h5-index 52.0

OBJECTIVE : Auditory attention in complex scenarios can be decoded by electroencephalography (EEG)-based cortical speech-envelope tracking. The relative root-mean-square (RMS) intensity is a valuable cue for the decomposition of speech into distinct characteristic segments. To improve auditory attention decoding (AAD) performance, this work proposed a novel segmented AAD approach to decode target speech envelopes from different RMS-level-based speech segments.

APPROACH : Speech was decomposed into higher- and lower-RMS-level speech segments with a threshold of -10 dB relative RMS level. A support vector machine classifier was designed to identify higher- and lower-RMS-level speech segments. Segmented computational models were developed with the classification results of higher- and lower-RMS-level speech segments. Speech envelopes were reconstructed based on segmented decoding models for either higher- or lower-RMS-level speech segments.

MAIN RESULTS : Higher- and lower-RMS-level speech segments in continuous sentences could be identified robustly with classification accuracies that approximated or exceeded 80% based on corresponding EEG signals at 6 dB, 3 dB, 0 dB, -3 dB and -6 dB signal-to-mask ratios (SMRs). Compared with unified AAD decoding methods, the proposed segmented AAD approach achieved more accurate results in the reconstruction of target speech envelopes and in the detection of attentional directions. Moreover, the proposed segmented decoding method had higher information transfer rates and shorter minimum expected switch times compared with the unified decoder.

SIGNIFICANCE : This study revealed that EEG signals may be used to classify higher- and lower-RMS-level-based speech segments across a wide range of SMR conditions (from 6 dB to -6 dB). A novel finding was that the specific information in different RMS-level-based speech segments facilitated EEG-based decoding of auditory attention. The significantly improved AAD accuracies and information transfer rates of the segmented decoding method suggests that this proposed computational model may be an effective method for the application of neuro-controlled brain-computer interfaces in complex auditory scenes.

Wang Lei, Wu Ed X, Chen Fei

2021-May-06

EEG, RMS-level-based speech segments, auditory attention decoding (AAD), machine learning, signal-to-mask ratio (SMR), support vector machine (SVM), temporal response function (TRF)

General General

Efficacy of machine learning techniques in predicting groundwater fluctuations in agro-ecological zones of India.

In The Science of the total environment

In the 21st century, groundwater depletion is posing a serious threat to humanity throughout the world, particularly in developing nations. India being the largest consumer of groundwater in the world, dwindling groundwater storage has emerged as a serious concern in recent years. Consequently, the judicious and efficient management of vital groundwater resources is one of the grand challenges in India. Groundwater modeling is a promising tool to develop sustainable management strategies for the efficient utilization of this treasured resource. This study demonstrates a pragmatic framework for predicting seasonal groundwater levels at a large scale using real-world data. Three relatively powerful Machine Learning (ML) techniques viz., ANFIS (Adaptive Neuro-Fuzzy Inference System), Deep Neural Network (DNN) and Support Vector Machine (SVM) were employed for predicting seasonal groundwater levels at the country scale using in situ groundwater-level and pertinent meteorological data of 1996-2016. ANFIS, DNN and SVM models were developed for 18 Agro-Ecological Zones (AEZs) of India and their efficacy was evaluated using suitable statistical and graphical indicators. The findings of this study revealed that the DNN model is the most proficient in predicting seasonal groundwater levels in most AEZs, followed by the ANFIS model. However, the prediction ability of the three models is 'moderate' to 'very poor' in 3 AEZs ['Western Plain and Kutch Peninsula' in Western India, and 'Deccan Plateau (Arid)' and 'Eastern Ghats and Deccan Plateau' in Southern India]. It is recommended that groundwater-monitoring network and data acquisition systems be strengthened in India in order to ensure efficient use of modeling techniques for the sustainable management of groundwater resources.

Mohapatra Janaki B, Jha Piyush, Jha Madan K, Biswal Sabinaya

2021-Apr-24

ANFIS, Agro-Ecological Zone, DNN, Data-driven modeling, Groundwater-level prediction, Machine learning/artificial intelligence techniques, SVM

Public Health Public Health

Prenatal exposure to ambient fine particulate matter and early childhood neurodevelopment: A population-based birth cohort study.

In The Science of the total environment

Although previous studies have reported the adverse effect of air pollution exposure during pregnancy on neurodevelopment in children, epidemiological evidence is limited, and the results are inconsistent. This study aimed to explore the association between prenatal ambient fine particulate matter (PM2.5) exposure and early childhood neurodevelopment in a large birth cohort study of 4009 maternal-child pairs. Prenatal daily PM2.5 exposure concentrations at 1 km spatial revolution were estimated using high-performance machine-learning models. Neurodevelopmental outcomes of children at ages 2, 6, 12, and 24 months were assessed using the Ages and Stages Questionnaire (ASQ). Distributed lag nonlinear models were used to identify critical windows of prenatal PM2.5 exposure. General linear mixed models with binomially distributed errors were used to estimate the effect of prenatal PM2.5 exposure on suspected developmental delay (SDD) in five developmental domains based on the longitudinal design. Prenatal PM2.5 exposure was significantly associated with decreased scores for all neurodevelopmental domains of children at ages 2, 6, and 24 months. Each 10-μg/m3 increase in PM2.5 exposure was significantly associated with increased risk of SDD for all subjects (RR: 1.52 95% CI: 1.19, 2.03), specifically, in problem-solving domain for girls (RR: 2.23, 95% CI: 1.22, 4.35). Prenatal PM2.5 exposure in weeks 18 to 34 was significantly associated with both ASQ scores and SDDs. Our study proposed that prenatal PM2.5 exposure affected early childhood neurodevelopment evaluated with the ASQ scale. PM2.5 exposure might increase the risk of SDD for boys and girls, specifically in the problem-solving domain for girls.

Wang Pengpeng, Zhao Yingya, Li Jialin, Zhou Yuhan, Luo Ranran, Meng Xia, Zhang Yunhui

2021-Apr-24

Birth cohort, Critical windows, Fine particulate matter, Neurodevelopment

Surgery Surgery

Creation of an artificial intelligence model for intubation difficulty classification by deep learning (convolutional neural network) using face images: an observational study.

In Journal of intensive care ; h5-index 30.0

BACKGROUND : Tracheal intubation is the gold standard for securing the airway, and it is not uncommon to encounter intubation difficulties in intensive care units and emergency rooms. Currently, there is a need for an objective measure to assess intubation difficulties in emergency situations by physicians, residents, and paramedics who are unfamiliar with tracheal intubation. Artificial intelligence (AI) is currently used in medical imaging owing to advanced performance. We aimed to create an AI model to classify intubation difficulties from the patient's facial image using a convolutional neural network (CNN), which links the facial image with the actual difficulty of intubation.

METHODS : Patients scheduled for surgery at Yamagata University Hospital between April and August 2020 were enrolled. Patients who underwent surgery with altered facial appearance, surgery with altered range of motion in the neck, or intubation performed by a physician with less than 3 years of anesthesia experience were excluded. Sixteen different facial images were obtained from the patients since the day after surgery. All images were judged as "Easy"/"Difficult" by an anesthesiologist, and an AI classification model was created using deep learning by linking the patient's facial image and the intubation difficulty. Receiver operating characteristic curves of actual intubation difficulty and AI model were developed, and sensitivity, specificity, and area under the curve (AUC) were calculated; median AUC was used as the result. Class activation heat maps were used to visualize how the AI model classifies intubation difficulties.

RESULTS : The best AI model for classifying intubation difficulties from 16 different images was generated in the supine-side-closed mouth-base position. The accuracy was 80.5%; sensitivity, 81.8%; specificity, 83.3%; AUC, 0.864; and 95% confidence interval, [0.731-0.969], indicating that the class activation heat map was concentrated around the neck regardless of the background; the AI model recognized facial contours and identified intubation difficulties.

CONCLUSION : This is the first study to apply deep learning (CNN) to classify intubation difficulties using an AI model. We could create an AI model with an AUC of 0.864. Our AI model may be useful for tracheal intubation performed by inexperienced medical staff in emergency situations or under general anesthesia.

Hayasaka Tatsuya, Kawano Kazuharu, Kurihara Kazuki, Suzuki Hiroto, Nakane Masaki, Kawamae Kaneyuki

2021-May-06

AI, Activation heat map, Intubation difficulty, Tracheal intubation

General General

Preliminary results of a clinical research and innovation scholarship to prepare medical students to lead innovations in health care.

In Healthcare (Amsterdam, Netherlands)

There is consensus amongst national organizations to integrate health innovation and augmented intelligence (AI) into medical education. However, there is scant evidence to guide policymakers and medical educators working to revise curricula. This study presents academic, operational, and domain understanding outcomes for the first three cohorts of participants in a clinical research and innovation scholarship program.

Sendak Mark P, Gao Michael, Ratliff William, Whalen Krista, Nichols Marshall, Futoma Joseph, Balu Suresh

2021-May-03

Health innovation, Machine learning, Medical education, Training

General General

Field determination of hazardous chemicals in public security by using a hand-held Raman spectrometer and a deep architecture-search network.

In Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy

With the advanced development of miniaturization and integration of instruments, Raman spectroscopy (RS) has demonstrated its great significance because of its non-invasive property and fingerprint identification ability, and extended its applications in public security, especially for hazardous chemicals. However, the fast and accurate RS analysis of hazardous chemicals in field test by non-professionals is still challenging due to the lack of an effective and timely spectral-based chemical-discriminating solution. In this study, a platform was developed for the field determination of hazardous chemicals in public security by using a hand-held Raman spectrometer and a deep architecture-search network (DASN) incorporated into a cloud server. With the Raman spectra of 300 chemicals, DASN stands out with identification accuracy of 100% and outweighs other machine learning and deep learning methods. The network feature maps for the spectra of methamphetamine and ketamine focus on the main peaks of 1001 and 652 cm-1, which indicates the powerful feature extraction capability of DASN. Its receiver operating characteristic (ROC) curve completely encloses the other models, and the area under the curve is up to 1, implying excellent robustness. With the well-built platform combining RS, DASN, and cloud server, one test process including Raman measurement and identification can be performed in tens of seconds. Hence, the developed platform is simple, fast, accurate, and could be considered as a promising tool for hazardous chemical identification in public security on the scene.

Dong Ronglu, Wang Jinghong, Weng Shizhuang, Yuan Hecai, Yang Liangbao

2021-Apr-24

Deep architecture-search network, Deep learning, Hazardous chemical, Raman spectroscopy, Substance identification

General General

Learning emotions latent representation with CVAE for text-driven expressive audiovisual speech synthesis.

In Neural networks : the official journal of the International Neural Network Society

Great improvement has been made in the field of expressive audiovisual Text-to-Speech synthesis (EAVTTS) thanks to deep learning techniques. However, generating realistic speech is still an open issue and researchers in this area have been focusing lately on controlling the speech variability. In this paper, we use different neural architectures to synthesize emotional speech. We study the application of unsupervised learning techniques for emotional speech modeling as well as methods for restructuring emotions representation to make it continuous and more flexible. This manipulation of the emotional representation should allow us to generate new styles of speech by mixing emotions. We first present our expressive audiovisual corpus. We validate the emotional content of this corpus with three perceptual experiments using acoustic only, visual only and audiovisual stimuli. After that, we analyze the performance of a fully connected neural network in learning characteristics specific to different emotions for the phone duration aspect and the acoustic and visual modalities. We also study the contribution of a joint and separate training of the acoustic and visual modalities in the quality of the generated synthetic speech. In the second part of this paper, we use a conditional variational auto-encoder (CVAE) architecture to learn a latent representation of emotions. We applied this method in an unsupervised manner to generate features of expressive speech. We used a probabilistic metric to compute the overlapping degree between emotions latent clusters to choose the best parameters for the CVAE. By manipulating the latent vectors, we were able to generate nuances of a given emotion and to generate new emotions that do not exist in our database. For these new emotions, we obtain a coherent articulation. We conducted four perceptual experiments to evaluate our findings.

Dahmani Sara, Colotte Vincent, Girard Valérian, Ouni Slim

2021-Apr-21

Bidirectional long short-term memory (BLSTM), Conditional variational auto-encoder, Deep learning, Expressive audiovisual speech synthesis, Expressive talking avatar

General General

A simplified cluster model and a tool adapted for collaborative labeling of lung cancer CT scans.

In Computer methods and programs in biomedicine

BACKGROUND AND OBJECTIVE : Lung cancer is the most common type of cancer with a high mortality rate. Early detection using medical imaging is critically important for the long-term survival of the patients. Computer-aided diagnosis (CAD) tools can potentially reduce the number of incorrect interpretations of medical image data by radiologists. Datasets with adequate sample size, annotation, and truth are the dominant factors in developing and training effective CAD algorithms. The objective of this study was to produce a practical approach and a tool for the creation of medical image datasets.

METHODS : The proposed model uses the modified maximum transverse diameter approach to mark a putative lung nodule. The modification involves the possibility to use a set of overlapping spheres of appropriate size to approximate the shape of the nodule. The algorithm embedded in the model also groups the marks made by different readers for the same lesion. We used the data of 536 randomly selected patients of Moscow outpatient clinics to create a dataset of standard-dose chest computed tomography (CT) scans utilizing the double-reading approach with arbitration. Six volunteer radiologists independently produced a report for each scan using the proposed model with the main focus on the detection of lesions with sizes ranging from 3 to 30 mm. After this, an arbitrator reviewed their marks and annotations.

RESULTS : The maximum transverse diameter approach outperformed the alternative methods (3D box, ellipsoid, and complete outline construction) in a study of 10,000 computer-generated tumor models of different shapes in terms of accuracy and speed of nodule shape approximation. The markup and annotation of the CTLungCa-500 dataset revealed 72 studies containing no lung nodules. The remaining 464 CT scans contained 3151 lesions marked by at least one radiologist: 56%, 14%, and 29% of the lesions were malignant, benign, and non-nodular, respectively. 2887 lesions have the target size of 3-30 mm. Only 70 nodules were uniformly identified by all the six readers. An increase in the number of independent readers providing CT scans interpretations led to an accuracy increase associated with a decrease in agreement. The dataset markup process took three working weeks.

CONCLUSIONS : The developed cluster model simplifies the collaborative and crowdsourced creation of image repositories and makes it time-efficient. Our proof-of-concept dataset provides a valuable source of annotated medical imaging data for training CAD algorithms aimed at early detection of lung nodules. The tool and the dataset are publicly available at https://github.com/Center-of-Diagnostics-and-Telemedicine/FAnTom.git and https://mosmed.ai/en/datasets/ct_lungcancer_500/, respectively.

Morozov S P, Gombolevskiy V A, Elizarov A B, Gusev M A, Novik V P, Prokudaylo S B, Bardin A S, Popov E V, Ledikhova N V, Chernina V Y, Blokhin I A, Nikolaev A E, Reshetnikov R V, Vladzymyrskyy A V, Kulberg N S

2021-Apr-18

Computed tomography (CT), Computer-aided diagnosis (CAD), Crowdsourcing, Lung nodule, Machine learning (ML), Training dataset, Weak labeling

General General

An integrated deep learning model for motor intention recognition of multi-class EEG Signals in upper limb amputees.

In Computer methods and programs in biomedicine

BACKGROUND AND OBJECTIVE : Recognition of motor intention based on electroencephalogram (EEG) signals has attracted considerable research interest in the field of pattern recognition due to its notable application of non-muscular communication and control for those with severe motor disabilities. In analysis of EEG data, achieving a higher classification performance is dependent on the appropriate representation of EEG features which is mostly characterized by one unique frequency before applying a learning model. Neglecting other frequencies of EEG signals could deteriorate the recognition performance of the model because each frequency has its unique advantages. Motivated by this idea, we propose to obtain distinguishable features with different frequencies by introducing an integrated deep learning model to accurately classify multiple classes of upper limb movement intentions.

METHODS : The proposed model is a combination of long short-term memory (LSTM) and stacked autoencoder (SAE). To validate the method, four high-level amputees were recruited to perform five motor intention tasks. The acquired EEG signals were first preprocessed before exploring the consequence of input representation on the performance of LSTM-SAE by feeding four frequency bands related to the tasks into the model. The learning model was further improved by t-distributed stochastic neighbor embedding (t-SNE) to eliminate feature redundancy, and to enhance the motor intention recognition.

RESULTS : The experimental results of the classification performance showed that the proposed model achieves an average performance of 99.01% for accuracy, 99.10% for precision, 99.09% for recall, 99.09% for f1_score, 99.77% for specificity, and 99.0% for Cohen's kappa, across multi-subject and multi-class scenarios. Further evaluation with 2-dimensional t-SNE revealed that the signal decomposition has a distinct multi-class separability in the feature space.

CONCLUSION : This study demonstrated the predominance of the proposed model in its ability to accurately classify upper limb movements from multiple classes of EEG signals, and its potential application in the development of a more intuitive and naturalistic prosthetic control.

Idowu Oluwagbenga Paul, Ilesanmi Ademola Enitan, Li Xiangxin, Samuel Oluwarotimi Williams, Fang Peng, Li Guanglin

2021-Apr-21

Brain computer interface (BCI), Electroencephalography (EEG), Long short-term memory (LSTM), Motor intention (MI), Stacked Autoencoder (SAE), t-distributed stochastic neighbor embedding (t-SNE)

General General

White matter hyperintensities segmentation using the ensemble U-Net with multi-scale highlighting foregrounds.

In NeuroImage ; h5-index 117.0

White matter hyperintensities (WMHs) are abnormal signals within the white matter region on the human brain MRI and have been associated with aging processes, cognitive decline, and dementia. In the current study, we proposed a U-Net with multi-scale highlighting foregrounds (HF) for WMHs segmentation. Our method, U-Net with HF, is designed to improve the detection of the WMH voxels with partial volume effects. We evaluated the segmentation performance of the proposed approach using the Challenge training dataset. Then we assessed the clinical utility of the WMH volumes that were automatically computed using our method and the Alzheimer's Disease Neuroimaging Initiative database. We demonstrated that the U-Net with HF significantly improved the detection of the WMH voxels at the boundary of the WMHs or in small WMH clusters quantitatively and qualitatively. Up to date, the proposed method has achieved the best overall evaluation scores, the highest dice similarity index, and the best F1-score among 39 methods submitted on the WMH Segmentation Challenge that was initially hosted by MICCAI 2017 and is continuously accepting new challengers. The evaluation of the clinical utility showed that the WMH volume that was automatically computed using U-Net with HF was significantly associated with cognitive performance and improves the classification between cognitive normal and Alzheimer's disease subjects and between patients with mild cognitive impairment and those with Alzheimer's disease. The implementation of our proposed method is publicly available using Dockerhub (https://hub.docker.com/r/wmhchallenge/pgs).

Park Gilsoon, Hong Jinwoo, Duffy Ben A, Lee Jong-Min, Kim Hosung

2021-May-03

U-Net, White matter hyperintensities, deep learning, multi-scale highlighting foregrounds, segmentation

General General

A Major Depressive Disorder Classification Framework based on EEG Signals using Statistical, Spectral, Wavelet, Functional Connectivity, and Nonlinear Analysis.

In Journal of neuroscience methods

BACKGROUND : Major depressive disorder (MDD) is a prevalent mental illness that is diagnosed through questionnaire-based approaches; however, these methods may not lead to an accurate diagnosis. In this regard, many studies have focused on using electroencephalogram (EEG) signals and machine learning techniques to diagnose MDD.

NEW METHOD : This paper proposes a machine learning framework for MDD diagnosis, which uses different types of EEG-derived features. The features are extracted using statistical, spectral, wavelet, functional connectivity, and nonlinear analysis methods. The sequential backward feature selection (SBFS) algorithm is also employed to perform feature selection. Various classifier models are utilized to select the best one for the proposed framework.

RESULTS : The proposed method is validated with a public EEG dataset, including the EEG data of 34 MDD patients and 30 healthy subjects. The evaluation of the proposed framework is conducted using 10-fold cross-validation, providing the metrics such as accuracy (AC), sensitivity (SE), specificity (SP), F1-score (F1), and false discovery rate (FDR). The best performance of the proposed method has provided an average AC of 99%, SE of 98.4%, SP of 99.6%, F1 of 98.9%, and FDR of 0.4% using the support vector machine with RBF kernel (RBFSVM) classifier.

COMPARISON WITH EXISTING METHODS : The obtained results demonstrate that the proposed method outperforms other approaches for MDD classification based on EEG signals.

CONCLUSIONS : According to the obtained results, a highly accurate MDD diagnosis would be provided using the proposed method, while it can be utilized to develop a computer-aided diagnosis (CAD) tool for clinical purposes.

Movahed Reza Akbari, Jahromi Gila Pirzad, Shahyad Shima, Meftahi Gholam Hossein

2021-May-03

Computer-aided diagnosis(CAD), Depression, Electroencephalogram(EEG), Machine learning, Major depressive disorder(MDD)

General General

Deconstructing the mouse olfactory percept through an ethological atlas.

In Current biology : CB

Odor perception in non-humans is poorly understood. Here, we generated the most comprehensive mouse olfactory ethological atlas to date, consisting of behavioral responses to a diverse panel of 73 odorants, including 12 at multiple concentrations. These data revealed that mouse behavior is incredibly diverse and changes in response to odorant identity and concentration. Using only behavioral responses observed in other mice, we could predict which of two odorants was presented to a held-out mouse 82% of the time. Considering all 73 possible odorants, we could uniquely identify the target odorant from behavior on the first try 20% of the time and 46% within five attempts. Although mouse behavior is difficult to predict from human perception, they share three fundamental properties: first, odor valence parameters explained the highest variance of olfactory perception. Second, physicochemical properties of odorants can be used to predict the olfactory percept. Third, odorant concentration quantitatively and qualitatively impacts olfactory perception. These results increase our understanding of mouse olfactory behavior and how it compares to human odor perception and provide a template for future comparative studies of olfactory percepts among species.

Manoel Diogo, Makhlouf Melanie, Arayata Charles J, Sathappan Abbirami, Da’as Sahar, Abdelrahman Doua, Selvaraj Senthil, Hasnah Reem, Mainland Joel D, Gerkin Richard C, Saraiva Luis R

2021-Apr-29

behavior, chemoinformatics, machine learning, odor, olfactory, perception

General General

Pharmacists' perceptions of a machine learning model for the identification of atypical medication orders.

In Journal of the American Medical Informatics Association : JAMIA

OBJECTIVES : The study sought to assess the clinical performance of a machine learning model aiming to identify unusual medication orders.

MATERIALS AND METHODS : This prospective study was conducted at CHU Sainte-Justine, Canada, from April to August 2020. An unsupervised machine learning model based on GANomaly and 2 baselines were trained to learn medication order patterns from 10 years of data. Clinical pharmacists dichotomously (typical or atypical) labeled orders and pharmacological profiles (patients' medication lists). Confusion matrices, areas under the precision-recall curve (AUPRs), and F1 scores were calculated.

RESULTS : A total of 12 471 medication orders and 1356 profiles were labeled by 25 pharmacists. Medication order predictions showed a precision of 35%, recall (sensitivity) of 26%, and specificity of 97% as compared with pharmacist labels, with an AUPR of 0.25 and an F1 score of 0.30. Profile predictions showed a precision of 49%, recall of 75%, and specificity of 82%, with an AUPR of 0.60, and an F1 score of 0.59. The model performed better than the baselines. According to the pharmacists, the model was a useful screening tool, and 9 of 15 participants preferred predictions by medication, rather than by profile.

DISCUSSION : Predictions for profiles had higher F1 scores and recall compared with medication order predictions. Although the performance was much better for profile predictions, pharmacists generally preferred medication order predictions.

CONCLUSIONS : Based on the AUPR, this model showed better performance for the identification of atypical pharmacological profiles than for medication orders. Pharmacists considered the model a useful screening tool. Improving these predictions should be prioritized in future research to maximize clinical impact.

Hogue Sophie-Camille, Chen Flora, Brassard Geneviève, Lebel Denis, Bussières Jean-François, Durand Audrey, Thibault Maxime

2021-May-06

clinical, clinical pharmacy information systems, decision support systems, hospital pharmaceutical services, machine learning, medical order entry systems

Internal Medicine Internal Medicine

Clinical factors associated with rapid treatment of sepsis.

In PloS one ; h5-index 176.0

PURPOSE : To understand what clinical presenting features of sepsis patients are historically associated with rapid treatment involving antibiotics and fluids, as appropriate.

DESIGN : This was a retrospective, observational cohort study using a machine-learning model with an embedded feature selection mechanism (gradient boosting machine).

METHODS : For adult patients (age ≥ 18 years) who were admitted through Emergency Department (ED) meeting clinical criteria of severe sepsis from 11/2007 to 05/2018 at an urban tertiary academic medical center, we developed gradient boosting models (GBMs) using a total of 760 original and derived variables, including demographic variables, laboratory values, vital signs, infection diagnosis present on admission, and historical comorbidities. We identified the most impactful factors having strong association with rapid treatment, and further applied the Shapley Additive exPlanation (SHAP) values to examine the marginal effects for each factor.

RESULTS : For the subgroups with or without fluid bolus treatment component, the models achieved high accuracy of area-under-receiver-operating-curve of 0.91 [95% CI, 0.86-0.95] and 0.84 [95% CI, 0.81-0.86], and sensitivity of 0.81[95% CI, 0.72-0.87] and 0.91 [95% CI, 0.81-0.97], respectively. We identified the 20 most impactful factors associated with rapid treatment for each subgroup. In the non-hypotensive subgroup, initial physiological values were the most impactful to the model, while in the fluid bolus subgroup, value minima and maxima tended to be the most impactful.

CONCLUSION : These machine learning methods identified factors associated with rapid treatment of severe sepsis patients from a large volume of high-dimensional clinical data. The results provide insight into differences in the rapid provision of treatment among patients with sepsis.

Song Xing, Liu Mei, Waitman Lemuel R, Patel Anurag, Simpson Steven Q

2021

General General

Artificial intelligence and robotisation in the EU - should we change OHS law?

In Journal of occupational medicine and toxicology (London, England)

BACKGROUND : Technological progress in the twenty-first century offers real chances for economic development of the European Union (EU). The purpose of this publication is to analyse risks and threats relating to Occupational Health and Safety (OHS) considerations in the context of scientific and technological development. The article attempts the analysis of whether current legislation of the European Union enables good protection of workers' health in the performance of their duties using robots, artificial intelligence (AI). A feature of robotisation and AI may be new challenges in OHS protection. The analysis performed aims to determine whether threats posted by working with Artificial Intelligence are serious enough for the EU Legislator to focus on implementation of new OHS regulations.

METHODS : The analysis was carried out on the basis of current legal regulations related to the protection of employee's health in the European Union. The study used literature related to robotisation with artificial intelligence and health and safety at work in the working environment.

RESULTS : Given the new psychological and physical threats related to the use of AI robots, it is necessary to expand the EU legislation with general guidelines for the use of intelligent robots in the work environment. Indeed, such robots must be defined in the applicable legal framework. Employers should also define, as part of their internal regulations, the procedures for employee communication with artificial intelligence, and relevantly update their training in the OHS area.

CONCLUSIONS : The developments in AI-assisted robots come with inherent risks and threats to the working environment. New challenges create the need for adapting EU laws to changing reality. In order to structure European Union legislation on health and safety at work, these changes could be defined in a single piece of legislation covering robotics and AI after detailed analysis, dialogue, and debate.

Jarota Maciej

2021-May-05

Artificial intelligence, Employee health, Employer, European Union, European legislation

General General

Machine learning of EEG spectra classifies unconsciousness during GABAergic anesthesia.

In PloS one ; h5-index 176.0

In current anesthesiology practice, anesthesiologists infer the state of unconsciousness without directly monitoring the brain. Drug- and patient-specific electroencephalographic (EEG) signatures of anesthesia-induced unconsciousness have been identified previously. We applied machine learning approaches to construct classification models for real-time tracking of unconscious state during anesthesia-induced unconsciousness. We used cross-validation to select and train the best performing models using 33,159 2s segments of EEG data recorded from 7 healthy volunteers who received increasing infusions of propofol while responding to stimuli to directly assess unconsciousness. Cross-validated models of unconsciousness performed very well when tested on 13,929 2s EEG segments from 3 left-out volunteers collected under the same conditions (median volunteer AUCs 0.99-0.99). Models showed strong generalization when tested on a cohort of 27 surgical patients receiving solely propofol collected in a separate clinical dataset under different circumstances and using different hardware (median patient AUCs 0.95-0.98), with model predictions corresponding with actions taken by the anesthesiologist during the cases. Performance was also strong for 17 patients receiving sevoflurane (alone or in addition to propofol) (median AUCs 0.88-0.92). These results indicate that EEG spectral features can predict unconsciousness, even when tested on a different anesthetic that acts with a similar neural mechanism. With high performance predictions of unconsciousness, we can accurately monitor anesthetic state, and this approach may be used to engineer infusion pumps to intelligibly respond to patients' neural activity.

Abel John H, Badgeley Marcus A, Meschede-Krasa Benyamin, Schamberg Gabriel, Garwood Indie C, Lecamwasam Kimaya, Chakravarty Sourish, Zhou David W, Keating Matthew, Purdon Patrick L, Brown Emery N

2021

General General

Drug-target interaction prediction using multi-head self-attention and graph attention network.

In IEEE/ACM transactions on computational biology and bioinformatics

Identifying drug-target interactions (DTIs) is an important step in the process of new drug discovery and drug repositioning. Accurate predictions for DTIs can improve efficiency in drug discovery and development. Although rapid advances in deep learning technologies have generated various computational methods, it is still appealing to further investigate how to design efficient networks for predicting DTIs. In this study, we propose an end-to-end deep learning method (called MHSADTI) to predict DTIs based on the graph attention network and multi-head self-attention mechanism models. First, the characteristics of drugs and proteins are extracted by the graph attention network model and multi-head self-attention mechanism, respectively. Then, the attention scores are used to consider which amino acid subsequence in a protein is more important for the drug to predict its interactions. Finally, we predict DTIs by a fully connected layer network after obtaining the feature vectors of drugs and proteins. The experiments in four datasets, human, C.elegans, DUD-E and DrugBank show our method outperforms the state-of-the-art method in terms of AUC, Precision, Recall, Recall, AUPR and F1-score. In addition, the case studies further demonstrate that our method can provide effective visualization to interpret the prediction results from biological insights.

Cheng Zhongjian, Yan Cheng, Wu Fangxiang, Wang Jianxin

2021-May-06

General General

DRONE: Dual-domain Residual-based Optimization NEtwork for Sparse-view CT Reconstruction.

In IEEE transactions on medical imaging ; h5-index 74.0

Deep learning has attracted rapidly increasing attention in the field of tomographic image reconstruction, especially for CT, MRI, PET/SPECT, ultrasound and optical imaging. Among various topics, sparse-view CT remains a challenge which targets a decent image reconstruction from ultra-sparse projections. To address this challenge, in this article we propose a Dual-domain Residual-based Optimization NEtwork (DRONE). DRONE consists of three modules respectively for embedding, refinement, and awareness. In the embedding module, a sparse sinogram is first extended. Then, sparse-view artifacts are effectively suppressed by the image domain networks. After that, the refinement module focuses on the recovery of image details in the residual data and image domains synergistically. Finally, the results from embedding and refinement components in the data and image domains are regularized for optimized image quality in the awareness module, which ensures the consistency between measurements and images with the kernel awareness of compressed sensing. The DRONE network is trained, validated, and tested on preclinical and clinical datasets, demonstrating its merits in edge preservation, feature recovery, and reconstruction accuracy.

Wu Weiwen, Wu Weiwen, Hu Dianlin, Niu Chuang, Yu Hengyong, Vardhanabhuti Varut, Wang Ge

2021-May-06

Ophthalmology Ophthalmology

Reporting guidelines for artificial intelligence in healthcare research.

In Clinical & experimental ophthalmology

Reporting guidelines are structured tools developed using explicit methodology that specify the minimum information required by researchers when reporting a study. The use of AI reporting guidelines that address potential sources of bias specific to studies involving AI interventions has the potential to improve the quality of AI studies, through improvements in their design and delivery, and the completeness and transparency of their reporting. With a number of guidance documents relating to AI studies emerging from different specialist societies, this Review article provides researchers with some key principles for selecting the most appropriate reporting guidelines for a study involving an AI intervention. As the main determinants of a high-quality study are contained within the methodology of the study design rather than the intervention, researchers are recommended to use reporting guidelines that are specific to the study design, and then supplement them with AI-specific guidance contained within available AI reporting guidelines. This article is protected by copyright. All rights reserved.

Ibrahim Hussein, Liu Xiaoxuan, Denniston Alastair K

2021-May-06

Artificial intelligence, Checklist, Guidelines, Machine learning, Research design, Research report

Public Health Public Health

Economic Burden of Multiple Sclerosis in Low- and Middle-Income Countries: A Systematic Review.

In PharmacoEconomics

BACKGROUND : Although the economic burden of multiple sclerosis (MS) in high-income countries (HICs) has been extensively studied, information on the costs of MS in low- and middle-income countries (LMICs) remains scarce. Moreover, no review synthesizing and assessing the costs of MS in LMICs has yet been undertaken.

OBJECTIVE : Our objective was to systematically identify and review the cost of illness (COI) of MS in LMICs to critically appraise the methodologies used, compare cost estimates across countries and by level of disease severity, and examine cost drivers.

METHODS : We conducted a systematic literature search for original studies in English, French, and Dutch containing prevalence or incidence-based cost data of MS in LMICs. The search was conducted in MEDLINE (Ovid), PubMed, Embase (Ovid), Cochrane Library, National Health Service Economic Evaluation Database (NHS EED), Econlit, and CINAHL (EBSCO) on July 2020 without restrictions on publication date. Recommended and validated methods were used for data extraction and analysis to make the results of the COI studies comparable. Costs were adjusted to $US, year 2019 values, using the World Bank purchasing power parity and inflated using the consumer price index.

RESULTS : A total of 14 studies were identified, all of which were conducted in upper-middle-income economies. Eight studies used a bottom-up approach for costing, and six used a top-down approach. Four studies used a societal perspective. The total annual cost per patient ranged between $US463 and 58,616. Costs varied across studies and countries, mainly because of differences regarding the inclusion of costs of disease-modifying therapies (DMTs), the range of cost items included, the methodological choices such as approaches used to estimate healthcare resource consumption, and the inclusion of informal care and productivity losses. Characteristics and methodologies of the included studies varied considerably, especially regarding the perspective adopted, cost data specification, and reporting of costs per severity levels. The total costs increased with greater disease severity. The cost ratios between different levels of MS severity within studies were relatively stable; costs were around 1-1.5 times higher for moderate versus mild MS and about two times higher for severe versus mild MS. MS drug costs were the main cost driver for less severe MS, whereas the proportion of direct non-medical costs and indirect costs increased with greater disease severity.

CONCLUSION : MS places a huge economic burden on healthcare systems and societies in LMICs. Methodological differences and substantial variations in terms of absolute costs were found between studies, which made comparison of studies challenging. However, the cost ratios across different levels of MS severity were similar, making comparisons between studies by disease severity feasible. Cost drivers were mainly DMTs and relapse treatments, and this was consistent across studies. Yet, the distribution of cost components varied with disease severity.

Dahham Jalal, Rizk Rana, Kremer Ingrid, Evers Silvia M A A, Hiligsmann Mickaël

2021-May-06

Radiology Radiology

Proposing a deep learning-based method for improving the diagnostic certainty of pulmonary nodules in CT scan of chest.

In European radiology ; h5-index 62.0

OBJECTIVE : To compare the performance of a deep learning (DL)-based method for diagnosing pulmonary nodules compared with radiologists' diagnostic approach in computed tomography (CT) of the chest.

MATERIALS AND METHODS : A total of 150 pathologically confirmed pulmonary nodules (60% malignant) assessed and reported by radiologists were included. CT images were processed by the proposed DL-based method to generate the probability of malignancy (0-100%), and the nodules were divided into the groups of benign (0-39.9%), indeterminate (40.0-59.9%), and malignant (60.0-100%). Taking the pathological results as the gold standard, we compared the diagnostic performance of the proposed DL-based method with the radiologists' diagnostic approach using the McNemar-Bowker test.

RESULTS : There was a statistically significant difference between the diagnosis results of the proposed DL-based method and the radiologists' diagnostic approach (p < 0.001). Moreover, there was no statistically significant difference in the composition of the diagnosis results between the proposed DL-based method and the radiologists' diagnostic approach (all p > 0.05). The difference in diagnostic accuracy between the proposed DL-based method (70%) and radiologists' diagnostic performance (64%) was not statistically significant (p = 0.243).

CONCLUSIONS : The proposed DL-based method achieved an accuracy comparable with the radiologists' diagnostic approach in clinical practice. Furthermore, its advantage in improving diagnostic certainty may raise the radiologists' confidence in diagnosing pulmonary nodules and may help clinical management. Therefore, the proposed DL-based method showed great potential in a certain clinical application.

KEY POINTS : • Deep learning-based method for diagnosing the pulmonary nodules in computed tomography provides a higher diagnostic certainty.

Wang Ya-Wen, Wang Jian-Wei, Yang Shou-Xin, Qi Lin-Lin, Lin Hao-Liang, Zhou Zhen, Yu Yi-Zhou

2021-May-06

Deep learning, Diagnosis, computer-assisted, Solitary pulmonary nodule, Tomography, X-ray computed

Surgery Surgery

Development and Validation of a Deep Learning Model Using Convolutional Neural Networks to Identify Scaphoid Fractures in Radiographs.

In JAMA network open

Importance : Scaphoid fractures are the most common carpal fracture, but as many as 20% are not visible (ie, occult) in the initial injury radiograph; untreated scaphoid fractures can lead to degenerative wrist arthritis and debilitating pain, detrimentally affecting productivity and quality of life. Occult scaphoid fractures are among the primary causes of scaphoid nonunions, secondary to delayed diagnosis.

Objective : To develop and validate a deep convolutional neural network (DCNN) that can reliably detect both apparent and occult scaphoid fractures from radiographic images.

Design, Setting, and Participants : This diagnostic study used a radiographic data set compiled for all patients presenting to Chang Gung Memorial Hospital (Taipei, Taiwan) and Michigan Medicine (Ann Arbor) with possible scaphoid fractures between January 2001 and December 2019. This group was randomly split into training, validation, and test data sets. The images were passed through a detection model to crop around the scaphoid and were then used to train a DCNN model based on the EfficientNetB3 architecture to classify apparent and occult scaphoid fractures. Data analysis was conducted from January to October 2020.

Exposures : A DCNN trained to discriminate radiographs with normal and fractured scaphoids.

Main Outcomes and Measures : Area under the receiver operating characteristic curve (AUROC), sensitivity, and specificity. Fracture localization was assessed using gradient-weighted class activation mapping.

Results : Of the 11 838 included radiographs (4917 [41.5%] with scaphoid fracture; 6921 [58.5%] without scaphoid fracture), 8356 (70.6%) were used for training, 1177 (9.9%) for validation, and 2305 (19.5%) for testing. In the testing test, the first DCNN achieved an overall sensitivity and specificity of 87.1% (95% CI, 84.8%-89.2%) and 92.1% (95% CI, 90.6%-93.5%), respectively, with an AUROC of 0.955 in distinguishing scaphoid fractures from scaphoids without fracture. Gradient-weighted class activation mapping closely corresponded to visible fracture sites. The second DCNN achieved an overall sensitivity of 79.0% (95% CI, 70.6%-71.6%) and specificity of 71.6% (95% CI, 69.0%-74.1%) with an AUROC of 0.810 when examining negative cases from the first model. Two-stage examination identified 20 of 22 cases (90.9%) of occult fracture.

Conclusions and Relevance : In this study, DCNN models were trained to identify scaphoid fractures. This suggests that such models may be able to assist with radiographic detection of occult scaphoid fractures that are not visible to human observers and to reliably detect fractures of other small bones.

Yoon Alfred P, Lee Yi-Lun, Kane Robert L, Kuo Chang-Fu, Lin Chihung, Chung Kevin C

2021-May-03

Internal Medicine Internal Medicine

Acute Exacerbation of a Chronic Obstructive Pulmonary Disease Prediction System Using Wearable Device Data, Machine Learning, and Deep Learning: Development and Cohort Study.

In JMIR mHealth and uHealth

BACKGROUND : The World Health Organization has projected that by 2030, chronic obstructive pulmonary disease (COPD) will be the third-leading cause of mortality and the seventh-leading cause of morbidity worldwide. Acute exacerbations of chronic obstructive pulmonary disease (AECOPD) are associated with an accelerated decline in lung function, diminished quality of life, and higher mortality. Accurate early detection of acute exacerbations will enable early management and reduce mortality.

OBJECTIVE : The aim of this study was to develop a prediction system using lifestyle data, environmental factors, and patient symptoms for the early detection of AECOPD in the upcoming 7 days.

METHODS : This prospective study was performed at National Taiwan University Hospital. Patients with COPD that did not have a pacemaker and were not pregnant were invited for enrollment. Data on lifestyle, temperature, humidity, and fine particulate matter were collected using wearable devices (Fitbit Versa), a home air quality-sensing device (EDIMAX Airbox), and a smartphone app. AECOPD episodes were evaluated via standardized questionnaires. With these input features, we evaluated the prediction performance of machine learning models, including random forest, decision trees, k-nearest neighbor, linear discriminant analysis, and adaptive boosting, and a deep neural network model.

RESULTS : The continuous real-time monitoring of lifestyle and indoor environment factors was implemented by integrating home air quality-sensing devices, a smartphone app, and wearable devices. All data from 67 COPD patients were collected prospectively during a mean 4-month follow-up period, resulting in the detection of 25 AECOPD episodes. For 7-day AECOPD prediction, the proposed AECOPD predictive model achieved an accuracy of 92.1%, sensitivity of 94%, and specificity of 90.4%. Receiver operating characteristic curve analysis showed that the area under the curve of the model in predicting AECOPD was greater than 0.9. The most important variables in the model were daily steps walked, stairs climbed, and daily distance moved.

CONCLUSIONS : Using wearable devices, home air quality-sensing devices, a smartphone app, and supervised prediction algorithms, we achieved excellent power to predict whether a patient would experience AECOPD within the upcoming 7 days. The AECOPD prediction system provided an effective way to collect lifestyle and environmental data, and yielded reliable predictions of future AECOPD events. Compared with previous studies, we have comprehensively improved the performance of the AECOPD prediction model by adding objective lifestyle and environmental data. This model could yield more accurate prediction results for COPD patients than using only questionnaire data.

Wu Chia-Tung, Li Guo-Hung, Huang Chun-Ta, Cheng Yu-Chieh, Chen Chi-Hsien, Chien Jung-Yien, Kuo Ping-Hung, Kuo Lu-Cheng, Lai Feipei

2021-May-06

chronic obstructive pulmonary disease, clinical decision support systems, health risk assessment, wearable device

General General

Use of Machine Learning Algorithms to Predict the Understandability of Health Education Materials: Development and Evaluation Study.

In JMIR medical informatics ; h5-index 23.0

BACKGROUND : Improving the understandability of health information can significantly increase the cost-effectiveness and efficiency of health education programs for vulnerable populations. There is a pressing need to develop clinically informed computerized tools to enable rapid, reliable assessment of the linguistic understandability of specialized health and medical education resources. This paper fills a critical gap in current patient-oriented health resource development, which requires reliable and accurate evaluation instruments to increase the efficiency and cost-effectiveness of health education resource evaluation.

OBJECTIVE : We aimed to translate internationally endorsed clinical guidelines to machine learning algorithms to facilitate the evaluation of the understandability of health resources for international students at Australian universities.

METHODS : Based on international patient health resource assessment guidelines, we developed machine learning algorithms to predict the linguistic understandability of health texts for Australian college students (aged 25-30 years) from non-English speaking backgrounds. We compared extreme gradient boosting, random forest, neural networks, and C5.0 decision tree for automated health information understandability evaluation. The 5 machine learning models achieved statistically better results compared to the baseline logistic regression model. We also evaluated the impact of each linguistic feature on the performance of each of the 5 models.

RESULTS : We found that information evidentness, relevance to educational purposes, and logical sequence were consistently more important than numeracy skills and medical knowledge when assessing the linguistic understandability of health education resources for international tertiary students with adequate English skills (International English Language Testing System mean score 6.5) and high health literacy (mean 16.5 in the Short Assessment of Health Literacy-English test). Our results challenge the traditional views that lack of medical knowledge and numerical skills constituted the barriers to the understanding of health educational materials.

CONCLUSIONS : Machine learning algorithms were developed to predict health information understandability for international college students aged 25-30 years. Thirteen natural language features and 5 evaluation dimensions were identified and compared in terms of their impact on the performance of the models. Health information understandability varies according to the demographic profiles of the target readers, and for international tertiary students, improving health information evidentness, relevance, and logic is critical.

Ji Meng, Liu Yanmeng, Zhao Mengdan, Lyu Ziqing, Zhang Boren, Luo Xin, Li Yanlin, Zhong Yin

2021-May-06

PEMAT, health education, machine learning, patient-oriented, understandability evaluation

General General

An artificial neural network-pharmacokinetic (ANN-PK) model and its interpretation using Shapley additive explanations.

In CPT: pharmacometrics & systems pharmacology

We developed a method to apply artificial neural networks (ANN) for predicting time-series pharmacokinetics, and an interpretable the ANN-pharmacokinetic (ANN-PK) model which can explain the evidence of prediction by applying Shapley additive explanations (SHAP). A previous population pharmacokinetic (popPK) model of cyclosporine A was used as the comparison model. The patients' data were used for the ANN-PK model input, and the output by ANN was the clearance (CL). The estimated CL value from the ANN were substituted into the 1-compartment with 1-order absorption model, the concentrations were calculated, and the parameters of ANN were updated by the back-propagation method. Kernel SHAP was applied to the trained model and the SHAP value of each input was calculated. The root mean squared error for the popPK model and the ANN-PK model were 41.1 ng/mL and 31.0 ng/mL, respectively. The goodness of fit plots for the ANN-PK model represented more convergence to y = x compared with that for the popPK model, with good model performance for the ANN-PK model. The most influential factors on CL output were age and body weight from the evaluation using Kernel SHAP, and these factors were incorporated into the popPK model as the significant covariates of CL. The ANN-PK model could handle time-series data and showed higher prediction accuracy then the conventional popPK model, and the scientific validity for the model could be evaluated by applying SHAP.

Ogami Chika, Tsuji Yasuhiro, Seki Hiroto, Kawano Hideaki, To Hideto, Matsumoto Yoshiaki, Hosono Hiroyuki

2021-May-06

Artificial neural network, Machine learning, Pharmacokinetics, Shapley additive explanation

General General

High-throughput soybean seeds phenotyping with convolutional neural networks and transfer learning.

In Plant methods

BACKGROUND : Effective soybean seed phenotyping demands large-scale accurate quantities of morphological parameters. The traditional manual acquisition of soybean seed morphological phenotype information is error-prone, and time-consuming, which is not feasible for large-scale collection. The segmentation of individual soybean seed is the prerequisite step for obtaining phenotypic traits such as seed length and seed width. Nevertheless, traditional image-based methods for obtaining high-throughput soybean seed phenotype are not robust and practical. Although deep learning-based algorithms can achieve accurate training and strong generalization capabilities, it requires a large amount of ground truth data which is often the limitation step.

RESULTS : We showed a novel synthetic image generation and augmentation method based on domain randomization. We synthesized a plenty of labeled image dataset automatedly by our method to train instance segmentation network for high throughput soybean seeds segmentation. It can pronouncedly decrease the cost of manual annotation and facilitate the preparation of training dataset. And the convolutional neural network can be purely trained by our synthetic image dataset to achieve a good performance. In the process of training Mask R-CNN, we proposed a transfer learning method which can reduce the computing costs significantly by finetuning the pre-trained model weights. We demonstrated the robustness and generalization ability of our method by analyzing the result of synthetic test datasets with different resolution and the real-world soybean seeds test dataset.

CONCLUSION : The experimental results show that the proposed method realized the effective segmentation of individual soybean seed and the efficient calculation of the morphological parameters of each seed and it is practical to use this approach for high-throughput objects instance segmentation and high-throughput seeds phenotyping.

Yang Si, Zheng Lihua, He Peng, Wu Tingting, Sun Shi, Wang Minjuan

2021-May-05

Deep learning, High throughput, Instance segmentation, Mask R-CNN, Seed phenotyping

General General

The ecological niche of reported rabies cases in Canada is similar to Alaska.

In Zoonoses and public health

The ecology of rabies in the circumpolar North is still not well understood. We use machine learning, a geographic information system and data explicit in time and space obtained for reported rabies cases and predictors in Canada to develop an ecological niche model for the distribution of reported rabies cases in the American north (Alaska and Canada). The ecological niche model based on reported rabies cases in Canada predicted reported rabies cases in Alaska, suggesting a rather robust inference and even similar drivers on a continental scale. As found in Alaska, proximity to human infrastructure-specifically along the coast-was a strong predictor in the detection of rabies cases in Canada. Also, this finding highlights the need for a more systematic landscape sampling for rabies infection model predictions to better understand and tackle the ecology of this important zoonotic disease on a landscape scale at some distance from human infrastructure in wilderness areas.

Huettmann Falk, Hueffer Karsten

2021-May-06

Alaska, Canada, Machine Learning, Rabies, geographic information system

General General

Distinct effects of prematurity on MRI metrics of brain functional connectivity, activity, and structure: Univariate and multivariate analyses.

In Human brain mapping

Premature birth affects the developmental trajectory of the brain during a period of intense maturation with possible lifelong consequences. To better understand the effect of prematurity on brain structure and function, we performed blood-oxygen-level dependent (BOLD) and anatomical magnetic resonance imaging (MRI) at 40 weeks of postmenstrual age on 88 newborns with variable gestational age (GA) at birth and no evident radiological alterations. We extracted measures of resting-state functional connectivity and activity in a set of 90 cortical and subcortical brain regions through the evaluation of BOLD correlations between regions and of fractional amplitude of low-frequency fluctuation (fALFF) within regions, respectively. Anatomical information was acquired through the assessment of regional volumes. We performed univariate analyses on each metric to examine the association with GA at birth, the spatial distribution of the effects, and the consistency across metrics. Moreover, a data-driven multivariate analysis (i.e., Machine Learning) framework exploited the high dimensionality of the data to assess the sensitivity of each metric to the effect of premature birth. Prematurity was associated with bidirectional alterations of functional connectivity and regional volume and, to a lesser extent, of fALFF. Notably, the effects of prematurity on functional connectivity were spatially diffuse, mainly within cortical regions, whereas effects on regional volume and fALFF were more focal, involving subcortical structures. While the two analytical approaches delivered consistent results, the multivariate analysis was more sensitive in capturing the complex pattern of prematurity effects. Future studies might apply multivariate frameworks to identify premature infants at risk of a negative neurodevelopmental outcome.

Chiarelli Antonio M, Sestieri Carlo, Navarra Riccardo, Wise Richard G, Caulo Massimo

2021-May-06

amplitude of low-frequency fluctuations, functional connectivity, magnetic resonance imaging, multivariate analysis, prematurity, regional volume

General General

Current and future deep learning algorithms for MS/MS-based small molecule structure elucidation.

In Rapid communications in mass spectrometry : RCM

RATIONALE : Structure elucidation of small molecules has been one of the cornerstone applications of mass spectrometry for decades. Despite the increasing availability of software tools, structure elucidation from MS/MS data remains a challenging task, leaving many spectra unidentified. However, as an increasing number of reference MS/MS spectra are being curated at a repository scale and shared on public servers, there is an exciting opportunity to develop powerful new deep learning (DL) models for automated structure elucidation.

ARCHITECTURES : Recent early-stage DL frameworks mostly follow a "two-step approach" that translates MS/MS spectra to database structures after first predicting molecular descriptors. The related architectures could suffer from: 1) computational complexity because of the separate training of descriptor-specific classifiers, 2) the high dimensional nature of mass spectral data and information loss due data preprocessing, 3) low substructure coverage and class imbalance problem of predefined molecular fingerprints. Inspired by successful DL frameworks employed in drug discovery fields, we have conceptualized and designed hypothetical DL architectures to tackle the above issues. For 1), we recommend multitask learning to achieve better performance with fewer classifiers by grouping structurally related descriptors. For 2) and 3), we introduce feature engineering to extract condensed and higher-order information from spectra and structure data. For instance, encoding spectra with subtrees and pre-calculated spectral patterns add peak interactions to the model input. Encoding structures with graph convolutional networks incorporates connectivity within a molecule. The joint embedding of spectra and structures can enable simultaneous spectral library and molecular database search.

CONCLUSIONS : We believe that in principle, given enough training data, adapted DL architectures, optimal hyperparameters and computing power, DL frameworks can predict small molecule structures, completely or at least partially, from MS/MS spectra. However, their performance and general applicability should be fairly evaluated against classical machine learning frameworks.

Liu Youzhong, De Vijlder Thomas, Bittremieux Wout, Laukens Kris, Heyndrickx Wouter

2021-May-06

Ophthalmology Ophthalmology

Detecting Sensitive Mobility Features for Parkinson's Disease Stages Via Machine Learning.

In Movement disorders : official journal of the Movement Disorder Society

BACKGROUND : It is not clear how specific gait measures reflect disease severity across the disease spectrum in Parkinson's disease (PD).

OBJECTIVE : To identify the gait and mobility measures that are most sensitive and reflective of PD motor stages and determine the optimal sensor location in each disease stage.

METHODS : Cross-sectional wearable-sensor records were collected in 332 patients with PD (Hoehn and Yahr scale I-III) and 100 age-matched healthy controls. Sensors were adhered to the participant's lower back, bilateral ankles, and wrists. Study participants walked in a ~15-meter corridor for 1 minute under two walking conditions: (1) preferred, usual walking speed and (2) walking while engaging in a cognitive task (dual-task). A subgroup (n = 303, 67% PD) also performed the Timed Up and Go test. Multiple machine-learning feature selection and classification algorithms were applied to discriminate between controls and PD and between the different PD severity stages.

RESULTS : High discriminatory values were found between motor disease stages with mean sensitivity in the range 72%-83%, specificity 69%-80%, and area under the curve (AUC) 0.76-0.90. Measures from upper-limb sensors best discriminated controls from early PD, turning measures obtained from the trunk sensor were prominent in mid-stage PD, and stride timing and regularity were discriminative in more advanced stages.

CONCLUSIONS : Applying machine-learning to multiple, wearable-derived features reveals that different measures of gait and mobility are associated with and discriminate distinct stages of PD. These disparate feature sets can augment the objective monitoring of disease progression and may be useful for cohort selection and power analyses in clinical trials of PD. © 2021 International Parkinson and Movement Disorder Society.

Mirelman Anat, Ben Or Frank Mor, Melamed Michal, Granovsky Lena, Nieuwboer Alice, Rochester Lynn, Del Din Silvia, Avanzino Laura, Pelosin Elisa, Bloem Bastiaan R, Della Croce Ugo, Cereatti Andrea, Bonato Paolo, Camicioli Richard, Ellis Theresa, Hamilton Jamie L, Hass Chris J, Almeida Quincy J, Inbal Maidan, Thaler Avner, Shirvan Julia, Cedarbaum Jesse M, Giladi Nir, Hausdorff Jeffrey M

2021-May-06

“Parkinsons disease”, accelerometer, gait, machine learning, wearables

Public Health Public Health

Twitter Surveillance at the Intersection of the Triangulum.

In Nicotine & tobacco research : official journal of the Society for Research on Nicotine and Tobacco

INTRODUCTION : A holistic public health surveillance approach can help capture the public's tobacco and marijuana-related attitudes and behaviors. Using publicly available data from Twitter, this is one of the first studies to describe key topics of discussions related to each intersection (e-cigarette, combustible tobacco, and marijuana) of the Triangulum framework.

METHOD : Twitter posts (n=999,447) containing marijuana, e-cigarette and combustible tobacco terms were collected from January 1, 2018, to December 23, 2019. Posts to Twitter with co-occurring mentions of keywords associated with the Triangulum were defined as an intersection (e-cigarettes and combustible tobacco, combustible tobacco and marijuana, e-cigarettes and marijuana, and marijuana, e-cigarettes and combustible tobacco). Text classifiers and unsupervised machine learning was used to identify predominant topics in posts.

RESULTS : Product Features and Cartridges were commonly referenced at the intersection of e-cigarette and marijuana-related conversations. Blunts and Cigars and Drugs and Alcohol were commonly referenced at the intersection of combustible tobacco and marijuana-related discussions. Flavors and Health Risks were discussed at the intersection of e-cigarette and combustible-related conversations, while discussions about Illicit products and Health risks were key topics of discussion when e-cigarettes, combustible tobacco, and marijuana were referenced all together in a single post.

CONCLUSION : By examining intersections of marijuana and tobacco products, this study offers inputs for designing comprehensive FDA regulations including regulating product features associated with appeal, improving enforcement to curb sales of illicit products, and informing the FDA's product review and standards procedures for tobacco products that can be used with marijuana.

IMPLICATIONS : This study is the first to leverage the Triangulum framework and Twitter data to describe key topics of discussions at the intersection of e-cigarette, combustible tobacco, and marijuana. Real-time health communication interventions can identify Twitter users posting in the context of e-cigarettes, combustible tobacco, and marijuana by automated methods and deliver tailored messages. This study also demonstrates the utility of Twitter data for surveillance of complex and evolving health behaviors.

Majmundar Anuja, Allem Jon-Patrick, Cruz Tess Boley, Unger Jennifer B, Pentz Mary Ann

2021-May-06

Internal Medicine Internal Medicine

A Novel 5-Cytokine Panel Outperforms Conventional Predictive Markers of Persistent Organ Failure in Acute Pancreatitis.

In Clinical and translational gastroenterology

INTRODUCTION : Existing laboratory markers and clinical scoring systems have shown suboptimal accuracies for early prediction of persistent organ failure (POF) in acute pancreatitis (AP). We used information theory and machine learning to select the best-performing panel of circulating cytokines for predicting POF early in the disease course and performed verification of the cytokine panel's prognostic accuracy in an independent AP cohort.

METHODS : The derivation cohort included 60 subjects with AP with early serum samples collected between 2007 and 2010. Twenty-five cytokines associated with an acute inflammatory response were ranked by computing the mutual information between their levels and the outcome of POF; 5 high-ranking cytokines were selected. These cytokines were subsequently measured in early serum samples of an independent prospective verification cohort of 133 patients (2012-2016), and the results were trained in a Random Forest classifier. Cross-validated performance metrics were compared with the predictive accuracies of conventional laboratory tests and clinical scores.

RESULTS : Angiopoietin 2, hepatocyte growth factor, interleukin 8, resistin, and soluble tumor necrosis factor receptor 1A were the highest-ranking cytokines in the derivation cohort; each reflects a pathologic process relevant to POF. A Random Forest classifier trained the cytokine panel in the verification cohort and achieved a 10-fold cross-validated accuracy of 0.89 (area under the curve 0.91, positive predictive value 0.89, and negative predictive value 0.90), which outperformed individual cytokines, laboratory tests, and clinical scores (all P ≤ 0.006).

DISCUSSION : We developed a 5-cytokine panel, which accurately predicts POF early in the disease process and significantly outperforms the prognostic accuracy of existing laboratory tests and clinical scores.

Langmead Christopher, Lee Peter J, Paragomi Pedram, Greer Phil, Stello Kim, Hart Phil A, Whitcomb David C, Papachristou Georgios I

2021-May-06

General General

An artificial intelligence algorithm for analyzing acetaminophen-associated toxic hepatitis.

In Human & experimental toxicology

INTRODUCTION : Very little artificial intelligence (AI) work has been performed to investigate acetaminophen-associated hepatotoxicity. The objective of this study was to develop an AI algorithm for analyzing weighted features for toxic hepatitis after acetaminophen poisoning.

METHODS : The medical records of 187 patients with acetaminophen poisoning treated at Chang Gung Memorial Hospital were reviewed. Patients were sorted into two groups according to their status of toxic hepatitis. A total of 40 clinical and laboratory features recorded on the first day of admission were selected for algorithm development. The random forest classifier (RFC) and logistic regression (LR) were used for artificial intelligence algorithm development.

RESULTS : The RFC-based AI model achieved the following results: accuracy = 92.5 ± 2.6%; sensitivity = 100%; specificity = 60%; precision = 92.3 ± 3.4%; and F1 = 96.0 ± 1.8%. The area under the receiver operating characteristic curve (AUROC) was approximately 0.98. The LR-based AI model achieved the following results: accuracy = 92.00 ± 2.9%; sensitivity = 100%; specificity = 20%; precision = 92.8 ± 3.4%; recall = 98.8 ± 3.4%; and F1 = 95.6 ± 1.5%. The AUROC was approximately 0.68. The weighted features were calculated, and the 10 most important weighted features for toxic hepatitis were aspartate aminotransferase (ALT), prothrombin time, alanine aminotransferase (AST), time to hospital, platelet count, lymphocyte count, albumin, total bilirubin, body temperature and acetaminophen level.

CONCLUSION : The top five weighted features for acetaminophen-associated toxic hepatitis were ALT, prothrombin time, AST, time to hospital and platelet count.

Yen J-S, Hu C-C, Huang W-H, Hsu C-W, Yen T-H, Weng C-H

2021-May-06

Artificial intelligence, acetaminophen, machine learning, poisoning, toxic hepatitis

oncology Oncology

High tumor cell platelet-derived growth factor receptor beta expression is associated with shorter survival in malignant pleural epithelioid mesothelioma.

In The journal of pathology. Clinical research

Malignant pleural mesothelioma (MPM) has a rich stromal component containing mesenchymal fibroblasts. However, the properties and interplay of MPM tumor cells and their surrounding stromal fibroblasts are poorly characterized. Our objective was to spatially profile known mesenchymal markers in both tumor cells and associated fibroblasts and correlate their expression with patient survival. The primary study cohort consisted of 74 MPM patients, including 16 patients who survived at least 60 months. We analyzed location-specific tissue expression of seven fibroblast markers in clinical samples using multiplexed fluorescence immunohistochemistry (mfIHC) and digital image analysis. Effect on survival was assessed using Cox regression analyses. The outcome measurement was all-cause mortality. Univariate analysis revealed that high expression of secreted protein acidic and cysteine rich (SPARC) and fibroblast activation protein in stromal cells was associated with shorter survival. Importantly, high expression of platelet-derived growth factor receptor beta (PDGFRB) in tumor cells, but not in stromal cells, was associated with shorter survival (hazard ratio [HR] = 1.02, p < 0.001). A multivariable survival analysis adjusted for clinical parameters and stromal mfIHC markers revealed that tumor cell PDGFRB and stromal SPARC remained independently associated with survival (HR = 1.01, 95% confidence interval [CI] = 1.00-1.03 and HR = 1.05, 95% CI = 1.00-1.11, respectively). The prognostic effect of PDGFRB was validated with an artificial intelligence-based analysis method and further externally validated in another cohort of 117 MPM patients. In external validation, high tumor cell PDGFRB expression associated with shorter survival, especially in the epithelioid subtype. Our findings suggest PDGFRB and SPARC as potential markers for risk stratification and as targets for therapy.

Ollila Hely, Paajanen Juuso, Wolff Henrik, Ilonen Ilkka, Sutinen Eva, Välimäki Katja, Östman Arne, Anttila Sisko, Kettunen Eeva, Räsänen Jari, Kallioniemi Olli, Myllärniemi Marjukka, Mäyränpää Mikko I, Pellinen Teijo

2021-May-06

fibroblast, mesothelioma, platelet-derived growth factor receptor beta, prognosis

General General

Re-Interpreting Complex Atrial Tachycardia Maps Using Global Atrial Vectors.

In Journal of cardiovascular electrophysiology

The mapping and ablation of Atrial Tachycardias (AT) can be challenging, particularly in patients with prior ablation or structural atrial disease. This article is protected by copyright. All rights reserved.

Rodrigo Miguel, Narayan Sanjiv M

2021-May-05

Artificial intelligence, Atrial Fibrillation, Atrial tachycardia, Machine learning, Mapping

General General

Mapping ethico-legal principles for the use of artificial intelligence in gastroenterology.

In Journal of gastroenterology and hepatology ; h5-index 51.0

The rapid development of artificial intelligence (AI) and digital health raise concerns about equitable access to innovative interventions, appropriate use of health data and privacy, inclusiveness, bias and discrimination, and even changes to the clinician-patient relationship. This article outlines a number of ethical and legal issues when examining the use of AI in gastroenterology. Substantive ethico-legal principles including respect for persons, privacy and confidentiality, integrity, conflict of interest, beneficence, nonmaleficence, and justice, are discussed. Much of what we articulated is relevant to the use of AI in other medical fields. Going forward, consorted efforts should be use to address more particular and concrete problems, but for now, a principle-based approach is best used in problem-solving.

Stewart Cameron, Wong Stephen K Y, Sung Joseph J Y

2021-May

artificial intelligence, ethics, legal

General General

Tongue image quality assessment based on a deep convolutional neural network.

In BMC medical informatics and decision making ; h5-index 38.0

BACKGROUND : Tongue diagnosis is an important research field of TCM diagnostic technology modernization. The quality of tongue images is the basis for constructing a standard dataset in the field of tongue diagnosis. To establish a standard tongue image database in the TCM industry, we need to evaluate the quality of a massive number of tongue images and add qualified images to the database. Therefore, an automatic, efficient and accurate quality control model is of significance to the development of intelligent tongue diagnosis technology for TCM.

METHODS : Machine learning methods, including Support Vector Machine (SVM), Random Forest (RF), Gradient Boosting Decision Tree (GBDT), Adaptive Boosting Algorithm (Adaboost), Naïve Bayes, Decision Tree (DT), Residual Neural Network (ResNet), Convolution Neural Network developed by Visual Geometry Group at University of Oxford (VGG), and Densely Connected Convolutional Networks (DenseNet), were utilized to identify good-quality and poor-quality tongue images. Their performances were made comparisons by using metrics such as accuracy, precision, recall, and F1-Score.

RESULTS : The experimental results showed that the accuracy of the three deep learning models was more than 96%, and the accuracy of ResNet-152 and DenseNet-169 was more than 98%. The model ResNet-152 obtained accuracy of 99.04%, precision of 99.05%, recall of 99.04%, and F1-score of 99.05%. The performances were better than performances of other eight models. The eight models are VGG-16, DenseNet-169, SVM, RF, GBDT, Adaboost, Naïve Bayes, and DT. ResNet-152 was selected as quality-screening model for tongue IQA.

CONCLUSIONS : Our research findings demonstrate various CNN models in the decision-making process for the selection of tongue image quality assessment and indicate that applying deep learning methods, specifically deep CNNs, to evaluate poor-quality tongue images is feasible.

Jiang Tao, Hu Xiao-Juan, Yao Xing-Hua, Tu Li-Ping, Huang Jing-Bin, Ma Xu-Xiang, Cui Ji, Wu Qing-Feng, Xu Jia-Tuo

2021-May-05

Deep learning, DenseNet, Quality assessment, ResNet, Tongue diagnosis

Radiology Radiology

Advanced CT techniques for assessing hepatocellular carcinoma.

In La Radiologia medica

Hepatocellular carcinoma (HCC) is the sixth-most common cancer in the world, and hepatic dynamic CT studies are routinely performed for its evaluation. Ongoing studies are examining advanced imaging techniques that may yield better findings than are obtained with conventional hepatic dynamic CT scanning. Dual-energy CT-, perfusion CT-, and artificial intelligence-based methods can be used for the precise characterization of liver tumors, the quantification of treatment responses, and for predicting the overall survival rate of patients. In this review, the advantages and disadvantages of conventional hepatic dynamic CT imaging are reviewed and the general principles of dual-energy- and perfusion CT, and the clinical applications and limitations of these technologies are discussed with respect to HCC. Finally, we address the utility of artificial intelligence-based methods for diagnosing HCC.

Nakamura Yuko, Higaki Toru, Honda Yukiko, Tatsugami Fuminari, Tani Chihiro, Fukumoto Wataru, Narita Keigo, Kondo Shota, Akagi Motonori, Awai Kazuo

2021-May-05

Artificial intelligence, Dual-energy CT, Hepatocellular carcinoma, Perfusion CT, Tomography, X-ray computed

Radiology Radiology

Radiomics complements clinical, radiological, and technical features to assess local control of colorectal cancer lung metastases treated with radiofrequency ablation.

In European radiology ; h5-index 62.0

OBJECTIVES : Radiofrequency ablation (RFA) of lung metastases of colorectal origin can improve patient survival and quality of life. Our aim was to identify pre- and per-RFA features predicting local control of lung metastases following RFA.

METHODS : This case-control single-center retrospective study included 119 lung metastases treated with RFA in 48 patients (median age: 60 years). Clinical, technical, and radiological data before and on early CT scan (at 48 h) were retrieved. After CT scan preprocessing, 64 radiomics features were extracted from pre-RFA and early control CT scans. Log-rank tests were used to detect categorical variables correlating with post-RFA local tumor progression-free survival (LTPFS). Radiomics prognostic scores (RPS) were developed on reproducible radiomics features using Monte-Carlo cross-validated LASSO Cox regressions.

RESULTS : Twenty-six of 119 (21.8%) nodules demonstrated local progression (median delay: 11.2 months). In univariate analysis, four non-radiomics variables correlated with post-RFA-LTPFS: nodule size (> 15 mm, p < 0.001), chosen electrode (with difference between covered array and nodule diameter < 20 mm or non-expandable electrode, p = 0.03), per-RFA intra-alveolar hemorrhage (IAH, p = 0.002), and nodule location into the ablation zone (not seen or in contact with borders, p = 0.005). The highest prognostic performance was reached with the multivariate model including a RPS built on 4 radiomics features from pre-RFA and early revaluation CT scans (cross-validated concordance index= 0.74) in which this RPS remained an independent predictor (cross-validated HR = 3.49, 95% confidence interval = [1.76 - 6.96]).

CONCLUSIONS : Technical, radiological, and radiomics features of the lung metastases before RFA and of the ablation zone at 48 h can help discriminate nodules at risk of local progression that could benefit from complementary local procedure.

KEY POINTS : • The highest prognostic performance to predict post-RFA LTPFS was reached with a parsimonious model including a radiomics score built with 4 radiomics features. • Nodule size, difference between electrode diameter, use of non-expandable electrode, per-RFA hemorrhage, and a tumor not seen or in contact with the ablation zone borders at 48-h CT were correlated with post-RFA LTPFS.

Markich Romane, Palussière Jean, Catena Vittorio, Cazayus Maxime, Fonck Marianne, Bechade Dominique, Buy Xavier, Crombé Amandine

2021-May-05

Colorectal neoplasms, Lung neoplasms, Machine learning, Radiology, interventional, Tomography, X-ray computed

General General

Fluorescence microscopy datasets for training deep neural networks.

In GigaScience

BACKGROUND : Fluorescence microscopy is an important technique in many areas of biological research. Two factors that limit the usefulness and performance of fluorescence microscopy are photobleaching of fluorescent probes during imaging and, when imaging live cells, phototoxicity caused by light exposure. Recently developed methods in machine learning are able to greatly improve the signal-to-noise ratio of acquired images. This allows researchers to record images with much shorter exposure times, which in turn minimizes photobleaching and phototoxicity by reducing the dose of light reaching the sample.

FINDINGS : To use deep learning methods, a large amount of data is needed to train the underlying convolutional neural network. One way to do this involves use of pairs of fluorescence microscopy images acquired with long and short exposure times. We provide high-quality datasets that can be used to train and evaluate deep learning methods under development.

CONCLUSION : The availability of high-quality data is vital for training convolutional neural networks that are used in current machine learning approaches.

Hagen Guy M, Bendesky Justin, Machado Rosa, Nguyen Tram-Anh, Kumar Tanmay, Ventura Jonathan

2021-May-05

convolutional neural networks, deep learning, fluorescence microscopy

General General

NSL2CD: identifying potential circRNA-disease associations based on network embedding and subspace learning.

In Briefings in bioinformatics

Many studies have evidenced that circular RNAs (circRNAs) are important regulators in various pathological processes and play vital roles in many human diseases, which could serve as promising biomarkers for disease diagnosis, treatment and prognosis. However, the functions of most of circRNAs remain to be unraveled, and it is time-consuming and costly to uncover those relationships between circRNAs and diseases by conventional experimental methods. Thus, identifying candidate circRNAs for human diseases offers new opportunities to understand the functional properties of circRNAs and the pathogenesis of diseases. In this study, we propose a novel network embedding-based adaptive subspace learning method (NSL2CD) for predicting potential circRNA-disease associations and discovering those disease-related circRNA candidates. The proposed method first calculates disease similarities and circRNA similarities by fully utilizing different data sources and learns low-dimensional node representations with network embedding methods. Then, we adopt an adaptive subspace learning model to discover potential associations between circRNAs and diseases. Meanwhile, an integrated weighted graph regularization term is imposed to preserve local geometric structures of data spaces, and L1,2-norm constraint is also incorporated into the model to realize the smoothness and sparsity of projection matrices. The experiment results show that NSL2CD achieves comparable performance under different evaluation metrics, and case studies further confirm its ability to discover potential candidate circRNAs for human diseases.

Xiao Qiu, Fu Yu, Yang Yide, Dai Jianhua, Luo Jiawei

2021-May-05

circRNA–disease associations, circular RNAs (circRNAs), disease-associated circRNAs, network embedding, subspace learning

General General

Prediction of tumor purity from gene expression data using machine learning.

In Briefings in bioinformatics

MOTIVATION : Bulk tumor samples used for high-throughput molecular profiling are often an admixture of cancer cells and non-cancerous cells, which include immune and stromal cells. The mixed composition can confound the analysis and affect the biological interpretation of the results, and thus, accurate prediction of tumor purity is critical. Although several methods have been proposed to predict tumor purity using high-throughput molecular data, there has been no comprehensive study on machine learning-based methods for the estimation of tumor purity.

RESULTS : We applied various machine learning models to estimate tumor purity. Overall, the models predicted the tumor purity accurately and showed a high correlation with well-established gold standard methods. In addition, we identified a small group of genes and demonstrated that they could predict tumor purity well. Finally, we confirmed that these genes were mainly involved in the immune system.

AVAILABILITY : The machine learning models constructed for this study are available at https://github.com/BonilKoo/ML_purity.

Koo Bonil, Rhee Je-Keun

2021-May-05

cancer genomics, machine learning, regression, tumor purity

General General

Machine Learning Techniques for Classifying the Mutagenic Origins of Point Mutations.

In Genetics ; h5-index 66.0

Mutations contribute significantly to developing diversity in biological capabilities. Mutagenesis is an adaptive feature of normal development, e.g. generating diversity in immune cells... There is increasing interest in developing diagnostics that discriminate individual mutagenic mechanisms in a range of applications that include identifying population-specific mutagenesis and resolving distinct mutation signatures in cancer samples. Analyses for these applications assume that mutagenic mechanisms have a distinct relationship with neighboring bases that allows them to be distinguished. Direct support for this assumption is limited to a small number of simple cases, e.g., CpG hypermutability. We have evaluated whether the mechanistic origin of a point mutation can be resolved using only sequence context for a more complicated case. We contrasted single nucleotide variants originating from the multitude of mutagenic processes that normally operate in the mouse germline with those induced by the potent mutagen N-ethyl-N-nitrosourea (ENU). The considerable overlap in the mutation spectra of these two samples make this a challenging problem. Employing a new, robust log-linear modeling method, we demonstrate that neighboring bases contain information regarding point mutation direction that differs between the ENU-induced and spontaneous mutation variant classes. A logistic regression classifier exhibited strong performance at discriminating between the different mutation classes. Concordance between the feature set of the best classifier and information content analyses suggest our results can be generalized to other mutation classification problems. We conclude that machine learning can be used to build a practical classification tool to identify the mutation mechanism for individual genetic variants. Software implementing our approach is freely available under an open-source license.

Zhu Yicheng, Ong Cheng Soon, Huttley Gavin A

2020-May-01

bioinformatics, context dependent mutation, germline mutation, log-linear model, machine learning, mutagenesis, mutation spectrum, sequence motif analysis

General General

Knowledge-Based Biomedical Data Science.

In Annual review of biomedical data science

Knowledge-based biomedical data science involves the design and implementation of computer systems that act as if they knew about biomedicine. Such systems depend on formally represented knowledge in computer systems, often in the form of knowledge graphs. Here we survey recent progress in systems that use formally represented knowledge to address data science problems in both clinical and biological domains, as well as progress on approaches for creating knowledge graphs. Major themes include the relationships between knowledge graphs and machine learning, the use of natural language processing to construct knowledge graphs, and the expansion of novel knowledge-based approaches to clinical and biological domains.

Callahan Tiffany J, Tripodi Ignacio J, Pielke-Lombardo Harrison, Hunter Lawrence E

2020-Jul

Semantic Web, knowledge discovery, knowledge graph, knowledge graph embeddings, natural language processing, ontology

General General

Artificial intelligence approaches and mechanisms for big data analytics: a systematic study.

In PeerJ. Computer science

Recent advances in sensor networks and the Internet of Things (IoT) technologies have led to the gathering of an enormous scale of data. The exploration of such huge quantities of data needs more efficient methods with high analysis accuracy. Artificial Intelligence (AI) techniques such as machine learning and evolutionary algorithms able to provide more precise, faster, and scalable outcomes in big data analytics. Despite this interest, as far as we are aware there is not any complete survey of various artificial intelligence techniques for big data analytics. The present survey aims to study the research done on big data analytics using artificial intelligence techniques. The authors select related research papers using the Systematic Literature Review (SLR) method. Four groups are considered to investigate these mechanisms which are machine learning, knowledge-based and reasoning methods, decision-making algorithms, and search methods and optimization theory. A number of articles are investigated within each category. Furthermore, this survey denotes the strengths and weaknesses of the selected AI-driven big data analytics techniques and discusses the related parameters, comparing them in terms of scalability, efficiency, precision, and privacy. Furthermore, a number of important areas are provided to enhance the big data analytics mechanisms in the future.

Rahmani Amir Masoud, Azhir Elham, Ali Saqib, Mohammadi Mokhtar, Ahmed Omed Hassan, Yassin Ghafour Marwan, Hasan Ahmed Sarkar, Hosseinzadeh Mehdi

2021

Artificial intelligence, Big data, Machine learning, Methods, Systematic literature review

General General

Harvesting social media sentiment analysis to enhance stock market prediction using deep learning.

In PeerJ. Computer science

Information gathering has become an integral part of assessing people's behaviors and actions. The Internet is used as an online learning site for sharing and exchanging ideas. People can actively give their reviews and recommendations for variety of products and services using popular social sites and personal blogs. Social networking sites, including Twitter, Facebook, and Google+, are examples of the sites used to share opinion. The stock market (SM) is an essential area of the economy and plays a significant role in trade and industry development. Predicting SM movements is a well-known and area of interest to researchers. Social networking perfectly reflects the public's views of current affairs. Financial news stories are thought to have an impact on the return of stock trend prices and many data mining techniques are used address fluctuations in the SM. Machine learning can provide a more accurate and robust approach to handle SM-related predictions. We sought to identify how movements in a company's stock prices correlate with the expressed opinions (sentiments) of the public about that company. We designed and implemented a stock price prediction accuracy tool considering public sentiment apart from other parameters. The proposed algorithm considers public sentiment, opinions, news and historical stock prices to forecast future stock prices. Our experiments were performed using machine-learning and deep-learning methods including Support Vector Machine, MNB classifier, linear regression, Naïve Bayes and Long Short-Term Memory. Our results validate the success of the proposed methodology.

Mehta Pooja, Pandya Sharnil, Kotecha Ketan

2021

Deep learning, LSTM, Machine learning, Sentiment analysis, Stock prediction

General General

Cyber-attack method and perpetrator prediction using machine learning algorithms.

In PeerJ. Computer science

Cyber-attacks have become one of the biggest problems of the world. They cause serious financial damages to countries and people every day. The increase in cyber-attacks also brings along cyber-crime. The key factors in the fight against crime and criminals are identifying the perpetrators of cyber-crime and understanding the methods of attack. Detecting and avoiding cyber-attacks are difficult tasks. However, researchers have recently been solving these problems by developing security models and making predictions through artificial intelligence methods. A high number of methods of crime prediction are available in the literature. On the other hand, they suffer from a deficiency in predicting cyber-crime and cyber-attack methods. This problem can be tackled by identifying an attack and the perpetrator of such attack, using actual data. The data include the type of crime, gender of perpetrator, damage and methods of attack. The data can be acquired from the applications of the persons who were exposed to cyber-attacks to the forensic units. In this paper, we analyze cyber-crimes in two different models with machine-learning methods and predict the effect of the defined features on the detection of the cyber-attack method and the perpetrator. We used eight machine-learning methods in our approach and concluded that their accuracy ratios were close. The Support Vector Machine Linear was found out to be the most successful in the cyber-attack method, with an accuracy rate of 95.02%. In the first model, we could predict the types of attacks that the victims were likely to be exposed to with a high accuracy. The Logistic Regression was the leading method in detecting attackers with an accuracy rate of 65.42%. In the second model, we predicted whether the perpetrators could be identified by comparing their characteristics. Our results have revealed that the probability of cyber-attack decreases as the education and income level of victim increases. We believe that cyber-crime units will use the proposed model. It will also facilitate the detection of cyber-attacks and make the fight against these attacks easier and more effective.

Bilen Abdulkadir, Özer Ahmet Bedri

2021

Artificial intelligence, Crime prediction, Cyber attack-crimes, Data analysis, Machine learning, Security and privacy

General General

DeepNetBim: deep learning model for predicting HLA-epitope interactions based on network analysis by harnessing binding and immunogenicity information.

In BMC bioinformatics

BACKGROUND : Epitope prediction is a useful approach in cancer immunology and immunotherapy. Many computational methods, including machine learning and network analysis, have been developed quickly for such purposes. However, regarding clinical applications, the existing tools are insufficient because few of the predicted binding molecules are immunogenic. Hence, to develop more potent and effective vaccines, it is important to understand binding and immunogenic potential. Here, we observed that the interactive association constituted by human leukocyte antigen (HLA)-peptide pairs can be regarded as a network in which each HLA and peptide is taken as a node. We speculated whether this network could detect the essential interactive propensities embedded in HLA-peptide pairs. Thus, we developed a network-based deep learning method called DeepNetBim by harnessing binding and immunogenic information to predict HLA-peptide interactions.

RESULTS : Quantitative class I HLA-peptide binding data and qualitative immunogenic data (including data generated from T cell activation assays, major histocompatibility complex (MHC) binding assays and MHC ligand elution assays) were retrieved from the Immune Epitope Database database. The weighted HLA-peptide binding network and immunogenic network were integrated into a network-based deep learning algorithm constituted by a convolutional neural network and an attention mechanism. The results showed that the integration of network centrality metrics increased the power of both binding and immunogenicity predictions, while the new model significantly outperformed those that did not include network features and those with shuffled networks. Applied on benchmark and independent datasets, DeepNetBim achieved an AUC score of 93.74% in HLA-peptide binding prediction, outperforming 11 state-of-the-art relevant models. Furthermore, the performance enhancement of the combined model, which filtered out negative immunogenic predictions, was confirmed on neoantigen identification by an increase in both positive predictive value (PPV) and the proportion of neoantigen recognition.

CONCLUSIONS : We developed a network-based deep learning method called DeepNetBim as a pan-specific epitope prediction tool. It extracted the attributes of the network as new features from HLA-peptide binding and immunogenic models. We observed that not only did DeepNetBim binding model outperform other updated methods but the combination of our two models showed better performance. This indicates further applications in clinical practice.

Yang Xiaoyun, Zhao Liyuan, Wei Fang, Li Jing

2021-May-05

Deep learning, Network analysis, T cell epitope prediction

General General

Knowledge distillation in deep learning and its applications.

In PeerJ. Computer science

Deep learning based models are relatively large, and it is hard to deploy such models on resource-limited devices such as mobile phones and embedded devices. One possible solution is knowledge distillation whereby a smaller model (student model) is trained by utilizing the information from a larger model (teacher model). In this paper, we present an outlook of knowledge distillation techniques applied to deep learning models. To compare the performances of different techniques, we propose a new metric called distillation metric which compares different knowledge distillation solutions based on models' sizes and accuracy scores. Based on the survey, some interesting conclusions are drawn and presented in this paper including the current challenges and possible research directions.

Alkhulaifi Abdolmaged, Alsahli Fahad, Ahmad Irfan

2021

Deep learning, Knowledge distillation, Model compression, Student model, Teacher model, Transferring knowledge

General General

Syntactic- and morphology-based text augmentation framework for Arabic sentiment analysis.

In PeerJ. Computer science

Arabic language is a challenging language for automatic processing. This is due to several intrinsic reasons such as Arabic multi-dialects, ambiguous syntax, syntactical flexibility and diacritics. Machine learning and deep learning frameworks require big datasets for training to ensure accurate predictions. This leads to another challenge faced by researches using Arabic text; as Arabic textual datasets of high quality are still scarce. In this paper, an intelligent framework for expanding or augmenting Arabic sentences is presented. The sentences were initially labelled by human annotators for sentiment analysis. The novel approach presented in this work relies on the rich morphology of Arabic, synonymy lists, syntactical or grammatical rules, and negation rules to generate new sentences from the seed sentences with their proper labels. Most augmentation techniques target image or video data. This study is the first work to target text augmentation for Arabic language. Using this framework, we were able to increase the size of the initial seed datasets by 10 folds. Experiments that assess the impact of this augmentation on sentiment analysis showed a 42% average increase in accuracy, due to the reliability and the high quality of the rules used to build this framework.

Duwairi Rehab, Abushaqra Ftoon

2021

Arabic text, Morphology-based augmentation, Natural language processing, Sentiment analysis, Text augmentation

General General

Enhancing semantic belief function to handle decision conflicts in SoS using k-means clustering.

In PeerJ. Computer science

Background : The endeavouring to offer complex special functions from individual systems gave rise to what is known as the System of Systems (SoS). SoS co-integrating systems together while allowing for absorbing more systems in the future. SoS as an integrated system simplifies operations, reduces costs, and ensures efficiency. However, conflict may result while co-integrating systems, violating the main benefits of SoS. This paper is concerned with enhancing the time required to detect and solve such conflicts.

Methods : We adopted the k-means clustering technique to enhance the detection and solving of conflict resulting while co-integrating new systems into an existing SoS. Instead of dealing with SoS as a single entity, we partition it into clusters. Each cluster contains nearby systems according to pre-specified criteria. We can consider each cluster a Sub SoS (S-SoS). By doing so, the conflict that may arise while co-integrating new systems can be detected and solved in a shorter time. We propose the Smart Semantic Belief Function Clustered System of Systems (SSBFCSoS), which is an enhancement of the Ontology Belief Function System of Systems (OBFSoS).

Results : The proposed method proved the ability to rapidly detect and resolve conflicts. It showed the ability to accommodate more systems as well, therefore achieving the objectives of SoS. In order to test the applicability of the SSBFCSoS and compare its performance with other approaches, two datasets were employed. They are (Glest & StarCraft Brood War). With each dataset, 15 test cases were examined. We achieved, on average, 89% in solving the conflict compared to 77% for other approaches. Moreover, it showed an acceleration of up to proportionality over previous approaches for about 16% in solving conflicts as well. Besides, it reduced the frequency of the same conflicts by approximately 23% better than the other method, not only in the same cluster but even while combining different clusters.

Elsayed Eman K, Ahmed Ahmed Sharaf Eldin, Younes Hebatullah Rashed

2021

Clustering, Component Systems (CS), Conflict, System of Systems (SoS), k-means

General General

Stance detection with BERT embeddings for credibility analysis of information on social media.

In PeerJ. Computer science

The evolution of electronic media is a mixed blessing. Due to the easy access, low cost, and faster reach of the information, people search out and devour news from online social networks. In contrast, the increasing acceptance of social media reporting leads to the spread of fake news. This is a minacious problem that causes disputes and endangers the societal stability and harmony. Fake news spread has gained attention from researchers due to its vicious nature. proliferation of misinformation in all media, from the internet to cable news, paid advertising and local news outlets, has made it essential for people to identify the misinformation and sort through the facts. Researchers are trying to analyze the credibility of information and curtail false information on such platforms. Credibility is the believability of the piece of information at hand. Analyzing the credibility of fake news is challenging due to the intent of its creation and the polychromatic nature of the news. In this work, we propose a model for detecting fake news. Our method investigates the content of the news at the early stage i.e., when the news is published but is yet to be disseminated through social media. Our work interprets the content with automatic feature extraction and the relevance of the text pieces. In summary, we introduce stance as one of the features along with the content of the article and employ the pre-trained contextualized word embeddings BERT to obtain the state-of-art results for fake news detection. The experiment conducted on the real-world dataset indicates that our model outperforms the previous work and enables fake news detection with an accuracy of 95.32%.

Karande Hema, Walambe Rahee, Benjamin Victor, Kotecha Ketan, Raghu T S

2021

** BERT, Credibility, LSTM, Misinformation, Fake news, Stance detection**

General General

Deterministic clustering based compressive sensing scheme for fog-supported heterogeneous wireless sensor networks.

In PeerJ. Computer science

Data acquisition problem in large-scale distributed Wireless Sensor Networks (WSNs) is one of the main issues that hinder the evolution of Internet of Things (IoT) technology. Recently, combination of Compressive Sensing (CS) and routing protocols has attracted much attention. An open question in this approach is how to integrate these techniques effectively for specific tasks. In this paper, we introduce an effective deterministic clustering based CS scheme (DCCS) for fog-supported heterogeneous WSNs to handle the data acquisition problem. DCCS employs the concept of fog computing, reduces total overhead and computational cost needed to self-organize sensor network by using a simple approach, and then uses CS at each sensor node to minimize the overall energy expenditure and prolong the IoT network lifetime. Additionally, the proposed scheme includes an effective algorithm for CS reconstruction called Random Selection Matching Pursuit (RSMP) to enhance the recovery process at the base station (BS) side with a complete scenario using CS. RSMP adds random selection process during the forward step to give opportunity for more columns to be selected as an estimated solution in each iteration. The results of simulation prove that the proposed technique succeeds to minimize the overall network power expenditure, prolong the network lifetime and provide better performance in CS data reconstruction.

Osamy Walid, Aziz Ahmed, M Khedr Ahmed

2021

CS reconstruction algorithms, Compressive sensing, Fog network, IoT, WSNs

General General

Strategy to improve the accuracy of convolutional neural network architectures applied to digital image steganalysis in the spatial domain.

In PeerJ. Computer science

In recent years, Deep Learning techniques applied to steganalysis have surpassed the traditional two-stage approach by unifying feature extraction and classification in a single model, the Convolutional Neural Network (CNN). Several CNN architectures have been proposed to solve this task, improving steganographic images' detection accuracy, but it is unclear which computational elements are relevant. Here we present a strategy to improve accuracy, convergence, and stability during training. The strategy involves a preprocessing stage with Spatial Rich Models filters, Spatial Dropout, Absolute Value layer, and Batch Normalization. Using the strategy improves the performance of three steganalysis CNNs and two image classification CNNs by enhancing the accuracy from 2% up to 10% while reducing the training time to less than 6 h and improving the networks' stability.

Tabares-Soto Reinel, Arteaga-Arteaga Harold Brayan, Mora-Rubio Alejandro, Bravo-Ortíz Mario Alejandro, Arias-Garzón Daniel, Alzate Grisales Jesús Alejandro, Burbano Jacome Alejandro, Orozco-Arias Simon, Isaza Gustavo, Ramos Pollan Raul

2021

Convolutional neural network, Deep learning, Steganalysis, Strategy

General General

Classification model for accuracy and intrusion detection using machine learning approach.

In PeerJ. Computer science

In today's cyber world, the demand for the internet is increasing day by day, increasing the concern of network security. The aim of an Intrusion Detection System (IDS) is to provide approaches against many fast-growing network attacks (e.g., DDoS attack, Ransomware attack, Botnet attack, etc.), as it blocks the harmful activities occurring in the network system. In this work, three different classification machine learning algorithms-Naïve Bayes (NB), Support Vector Machine (SVM), and K-nearest neighbor (KNN)-were used to detect the accuracy and reducing the processing time of an algorithm on the UNSW-NB15 dataset and to find the best-suited algorithm which can efficiently learn the pattern of the suspicious network activities. The data gathered from the feature set comparison was then applied as input to IDS as data feeds to train the system for future intrusion behavior prediction and analysis using the best-fit algorithm chosen from the above three algorithms based on the performance metrics found. Also, the classification reports (Precision, Recall, and F1-score) and confusion matrix were generated and compared to finalize the support-validation status found throughout the testing phase of the model used in this approach.

Agarwal Arushi, Sharma Purushottam, Alshehri Mohammed, Mohamed Ahmed A, Alfarraj Osama

2021

Intrusion detection system, K-Nearest Neighbors (KNN), Naive Bayes (NB), Support vector machine (SVM), UNSWNB15 dataset

General General

Multi-level aspect based sentiment classification of Twitter data: using hybrid approach in deep learning.

In PeerJ. Computer science

Social media is a vital source to produce textual data, further utilized in various research fields. It has been considered an essential foundation for organizations to get valuable data to assess the users' thoughts and opinions on a specific topic. Text classification is a procedure to assign tags to predefined classes automatically based on their contents. The aspect-based sentiment analysis to classify the text is challenging. Every work related to sentiment analysis approached this issue as the current research usually discusses the document-level and overall sentence-level analysis rather than the particularities of the sentiments. This research aims to use Twitter data to perform a finer-grained sentiment analysis at aspect-level by considering explicit and implicit aspects. This study proposes a new Multi-level Hybrid Aspect-Based Sentiment Classification (MuLeHyABSC) approach by embedding a feature ranking process with an amendment of feature selection method for Twitter and sentiment classification comprising of Artificial Neural Network; Multi-Layer Perceptron (MLP) is used to attain improved results. In this study, different machine learning classification methods were also implemented, including Random Forest (RF), Support Vector Classifier (SVC), and seven more classifiers to compare with the proposed classification method. The implementation of the proposed hybrid method has shown better performance and the efficiency of the proposed system was validated on multiple Twitter datasets to manifest different domains. We achieved better results for all Twitter datasets used for the validation purpose of the proposed method with an accuracy of 78.99%, 84.09%, 80.38%, 82.37%, and 84.72%, respectively, compared to the baseline approaches. The proposed approach revealed that the new hybrid aspect-based text classification functionality is enhanced, and it outperformed the existing baseline methods for sentiment classification.

Janjua Sadaf Hussain, Siddiqui Ghazanfar Farooq, Sindhu Muddassar Azam, Rashid Umer

2021

Aspect-based sentiment classification, Feature extraction, Feature selection, Hybrid approach, Information gain, Multi-layer perception, Principal component analysis

General General

A real-time approach of diagnosing rice leaf disease using deep learning-based faster R-CNN framework.

In PeerJ. Computer science

The rice leaves related diseases often pose threats to the sustainable production of rice affecting many farmers around the world. Early diagnosis and appropriate remedy of the rice leaf infection is crucial in facilitating healthy growth of the rice plants to ensure adequate supply and food security to the rapidly increasing population. Therefore, machine-driven disease diagnosis systems could mitigate the limitations of the conventional methods for leaf disease diagnosis techniques that is often time-consuming, inaccurate, and expensive. Nowadays, computer-assisted rice leaf disease diagnosis systems are becoming very popular. However, several limitations ranging from strong image backgrounds, vague symptoms' edge, dissimilarity in the image capturing weather, lack of real field rice leaf image data, variation in symptoms from the same infection, multiple infections producing similar symptoms, and lack of efficient real-time system mar the efficacy of the system and its usage. To mitigate the aforesaid problems, a faster region-based convolutional neural network (Faster R-CNN) was employed for the real-time detection of rice leaf diseases in the present research. The Faster R-CNN algorithm introduces advanced RPN architecture that addresses the object location very precisely to generate candidate regions. The robustness of the Faster R-CNN model is enhanced by training the model with publicly available online and own real-field rice leaf datasets. The proposed deep-learning-based approach was observed to be effective in the automatic diagnosis of three discriminative rice leaf diseases including rice blast, brown spot, and hispa with an accuracy of 98.09%, 98.85%, and 99.17% respectively. Moreover, the model was able to identify a healthy rice leaf with an accuracy of 99.25%. The results obtained herein demonstrated that the Faster R-CNN model offers a high-performing rice leaf infection identification system that could diagnose the most common rice diseases more precisely in real-time.

Bari Bifta Sama, Islam Md Nahidul, Rashid Mamunur, Hasan Md Jahid, Razman Mohd Azraai Mohd, Musa Rabiu Muazu, Ab Nasir Ahmad Fakhri, P P Abdul Majeed Anwar

2021

Deep learning, Faster R-CNN, Image processing, Object detection, Rice reaf disease detection

General General

An ensemble approach based on transformation functions for natural gas price forecasting considering optimal time delays.

In PeerJ. Computer science

Natural gas, known as the cleanest fossil fuel, plays a vital role in the economies of producing and consuming countries. Understanding and tracking the drivers of natural gas prices are of significant interest to the many economic sectors. Hence, accurately forecasting the price is very important not only for providing an effective factor for implementing energy policy but also for playing an extremely significant role in government strategic planning. The purpose of this study is to provide an approach to forecast the natural gas price. First, optimal time delays are identified by a new approach based on the Euclidean Distance between input and target vectors. Then, wavelet decomposition has been implemented to reduce noise. Moreover, fuzzy transform with different membership functions has been used for modeling uncertainty in time series. The wavelet decomposition and fuzzy transform have been integrated into the preprocessing stage. An ensemble method is used for integrating the outputs of various neural networks. The results depict that the proposed preprocessing methods used in this paper cause to improve the accuracy of natural gas price forecasting and consider uncertainty in time series.

Saghi Faramarz, Jahangoshai Rezaee Mustafa

2021

Ensemble artificial neural networks, Machine learning, Natural gas price, Optimal time delays, Time series forecasting

General General

Application of machine learning in predicting hospital readmissions: a scoping review of the literature.

In BMC medical research methodology

BACKGROUND : Advances in machine learning (ML) provide great opportunities in the prediction of hospital readmission. This review synthesizes the literature on ML methods and their performance for predicting hospital readmission in the US.

METHODS : This review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Extension for Scoping Reviews (PRISMA-ScR) Statement. The extraction of items was also guided by the Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies (CHARMS). Electronic databases PUBMED, MEDLINE, and EMBASE were systematically searched from January 1, 2015, through December 10, 2019. The articles were imported into COVIDENCE online software for title/abstract screening and full-text eligibility. Observational studies using ML techniques for hospital readmissions among US patients were eligible for inclusion. Articles without a full text available in the English language were excluded. A qualitative synthesis included study characteristics, ML algorithms utilized, and model validation, and quantitative analysis assessed model performance. Model performances in terms of Area Under the Curve (AUC) were analyzed using R software. Quality in Prognosis Studies (QUIPS) tool was used to assess the quality of the reviewed studies.

RESULTS : Of 522 citations reviewed, 43 studies met the inclusion criteria. A majority of the studies used electronic health records (24, 56%), followed by population-based data sources (15, 35%) and administrative claims data (4, 9%). The most common algorithms were tree-based methods (23, 53%), neural network (NN) (14, 33%), regularized logistic regression (12, 28%), and support vector machine (SVM) (10, 23%). Most of these studies (37, 85%) were of high quality. A majority of these studies (28, 65%) reported ML algorithms with an AUC above 0.70. There was a range of variability within AUC reported by these studies with a median of 0.68 (IQR: 0.64-0.76; range: 0.50-0.90).

CONCLUSIONS : The ML algorithms involving tree-based methods, NN, regularized logistic regression, and SVM are commonly used to predict hospital readmission in the US. Further research is needed to compare the performance of ML algorithms for hospital readmission prediction.

Huang Yinan, Talwar Ashna, Chatterjee Satabdi, Aparasu Rajender R

2021-May-06

Hospital readmission, Machine learning, Prediction, Scoping review

General General

Sacral acceleration can predict whole-body kinetics and stride kinematics across running speeds.

In PeerJ

Background : Stress fractures are injuries caused by repetitive loading during activities such as running. The application of advanced analytical methods such as machine learning to data from multiple wearable sensors has allowed for predictions of biomechanical variables associated with running-related injuries like stress fractures. However, it is unclear if data from a single wearable sensor can accurately estimate variables that characterize external loading during running such as peak vertical ground reaction force (vGRF), vertical impulse, and ground contact time. Predicting these biomechanical variables with a single wearable sensor could allow researchers, clinicians, and coaches to longitudinally monitor biomechanical running-related injury risk factors without expensive force-measuring equipment.

Purpose : We quantified the accuracy of applying quantile regression forest (QRF) and linear regression (LR) models to sacral-mounted accelerometer data to predict peak vGRF, vertical impulse, and ground contact time across a range of running speeds.

Methods : Thirty-seven collegiate cross country runners (24 females, 13 males) ran on a force-measuring treadmill at 3.8-5.4 m/s while wearing an accelerometer clipped posteriorly to the waistband of their running shorts. We cross-validated QRF and LR models by training them on acceleration data, running speed, step frequency, and body mass as predictor variables. Trained models were then used to predict peak vGRF, vertical impulse, and contact time. We compared predicted values to those calculated from a force-measuring treadmill on a subset of data (n = 9) withheld during model training. We quantified prediction accuracy by calculating the root mean square error (RMSE) and mean absolute percentage error (MAPE).

Results : The QRF model predicted peak vGRF with a RMSE of 0.150 body weights (BW) and MAPE of 4.27  ±  2.85%, predicted vertical impulse with a RMSE of 0.004 BW*s and MAPE of 0.80  ±  0.91%, and predicted contact time with a RMSE of 0.011 s and MAPE of 4.68  ±  3.00%. The LR model predicted peak vGRF with a RMSE of 0.139 BW and MAPE of 4.04  ±  2.57%, predicted vertical impulse with a RMSE of 0.002 BW*s and MAPE of 0.50  ±  0.42%, and predicted contact time with a RMSE of 0.008 s and MAPE of 3.50  ±  2.27%. There were no statistically significant differences between QRF and LR model prediction MAPE for peak vGRF (p = 0.549) or vertical impulse (p = 0.073), but the LR model's MAPE for contact time was significantly lower than the QRF model's MAPE (p = 0.0497).

Conclusions : Our findings indicate that the QRF and LR models can accurately predict peak vGRF, vertical impulse, and contact time (MAPE < 5%) from a single sacral-mounted accelerometer across a range of running speeds. These findings may be beneficial for researchers, clinicians, or coaches seeking to monitor running-related injury risk factors without force-measuring equipment.

Alcantara Ryan S, Day Evan M, Hahn Michael E, Grabowski Alena M

2021

Biomechanics, Ground reaction force, Inertial measurement unit, Injury, Machine learning, Stress fracture

General General

scAnt-an open-source platform for the creation of 3D models of arthropods (and other small objects).

In PeerJ

We present scAnt, an open-source platform for the creation of digital 3D models of arthropods and small objects. scAnt consists of a scanner and a Graphical User Interface, and enables the automated generation of Extended Depth Of Field images from multiple perspectives. These images are then masked with a novel automatic routine which combines random forest-based edge-detection, adaptive thresholding and connected component labelling. The masked images can then be processed further with a photogrammetry software package of choice, including open-source options such as Meshroom, to create high-quality, textured 3D models. We demonstrate how these 3D models can be rigged to enable realistic digital specimen posing, and introduce a novel simple yet effective method to include semi-realistic representations of approximately planar and transparent structures such as wings. As a result of the exclusive reliance on generic hardware components, rapid prototyping and open-source software, scAnt costs only a fraction of available comparable systems. The resulting accessibility of scAnt will (i) drive the development of novel and powerful methods for machine learning-driven behavioural studies, leveraging synthetic data; (ii) increase accuracy in comparative morphometric studies as well as extend the available parameter space with area and volume measurements; (iii) inspire novel forms of outreach; and (iv) aid in the digitisation efforts currently underway in several major natural history collections.

Plum Fabian, Labonte David

2021

3D, Digitisation, Macro imaging, Morphometry, Photogrammetry, Zoology

General General

Enhancing deep-learning training for phase identification in powder X-ray diffractograms.

In IUCrJ

Within the domain of analyzing powder X-ray diffraction (XRD) scans, manual examination of the recorded data is still the most popular method, but it requires some expertise and is time consuming. The usual workflow for the phase-identification task involves software for searching databases of known compounds and matching lists of d spacings and related intensities to the measured data. Most automated approaches apply some iterative procedure for the search/match process but fail to be generally reliable yet without the manual validation step of an expert. Recent advances in the field of machine and deep learning have led to the development of algorithms for use with diffraction patterns and are producing promising results in some applications. A limitation, however, is that thousands of training samples are required for the model to achieve a reliable performance and not enough measured samples are available. Accordingly, a framework for the efficient generation of thousands of synthetic XRD scans is presented which considers typical effects in realistic measurements and thus simulates realistic patterns for the training of machine- or deep-learning models. The generated data set can be applied to any machine- or deep-learning structure as training data so that the models learn to analyze measured XRD data based on synthetic diffraction patterns. Consequently, we train a convolutional neural network with the simulated diffraction patterns for application with iron ores or cements compounds and prove robustness against varying unit-cell parameters, preferred orientation and crystallite size in synthetic, as well as measured, XRD scans.

Schuetzke Jan, Benedix Alexander, Mikut Ralf, Reischl Markus

2021-May-01

X-ray diffraction, computational modelling, convolutional neural networks, deep learning, multiphase, phase identification

General General

Support Vector Machine Weather Prediction Technology Based on the Improved Quantum Optimization Algorithm.

In Computational intelligence and neuroscience

Emotion recognition is a research hotspot in the field of artificial intelligence. If the human-computer interaction system can sense human emotion and express emotion, it will make the interaction between the robot and human more natural. In this paper, a multimodal emotion recognition model based on many-objective optimization algorithm is proposed for the first time. The model integrates voice information and facial information and can simultaneously optimize the accuracy and uniformity of recognition. This paper compares the emotion recognition algorithm based on many-objective algorithm optimization with the single-modal emotion recognition model proposed in this paper and the ISMS_ALA model proposed by recent related research. The experimental results show that compared with the single-mode emotion recognition, the proposed model has a great improvement in each evaluation index. At the same time, the accuracy of emotion recognition is 2.88% higher than that of the ISMS_ALA model. The experimental results show that the many-objective optimization algorithm can effectively improve the performance of the multimodal emotion recognition model.

Zhang Jinlei, Qiu Xue, Li Xiang, Huang Zhijie, Wu Mingqiu, Dong Yumin

2021

General General

Home Textile Pattern Emotion Labeling Using Deep Multi-View Feature Learning.

In Frontiers in psychology ; h5-index 92.0

Different home textile patterns have different emotional expressions. Emotion evaluation of home textile patterns can effectively improve the retrieval performance of home textile patterns based on semantics. It can not only help designers make full use of existing designs and stimulate creative inspiration but also help users select designs and products that are more in line with their needs. In this study, we develop a three-stage framework for home textile pattern emotion labeling based on artificial intelligence. To be specific, first of all, three kinds of aesthetic features, i.e., shape, texture, and salient region, are extracted from the original home textile patterns. Then, a CNN (convolutional neural network)-based deep feature extractor is constructed to extract deep features from the aesthetic features acquired in the previous stage. Finally, a novel multi-view classifier is designed to label home textile patterns that can automatically learn the weight of each view. The three-stage framework is evaluated by our data and the experimental results show its promising performance in home textile patterns labeling.

Yang Juan, Zhang Yuanpeng

2021

deep learning, emotion labeling, feature selection, home textile pattern, multi-view learning

General General

Polo: an open-source graphical user interface for crystallization screening.

In Journal of applied crystallography

Polo is a Python-based graphical user interface designed to streamline viewing and analysis of images to monitor crystal growth, with a specific target to enable users of the High-Throughput Crystallization Screening Center at Hauptman-Woodward Medical Research Institute (HWI) to efficiently inspect their crystallization experiments. Polo aims to increase efficiency, reducing time spent manually reviewing crystallization images, and to improve the potential of identifying positive crystallization conditions. Polo provides a streamlined one-click graphical interface for the Machine Recognition of Crystallization Outcomes (MARCO) convolutional neural network for automated image classification, as well as powerful tools to view and score crystallization images, to compare crystallization conditions, and to facilitate collaborative review of crystallization screening results. Crystallization images need not have been captured at HWI to utilize Polo's basic functionality. Polo is free to use and modify for both academic and commercial use under the terms of the copyleft GNU General Public License v3.0.

Holleman Ethan T, Duguid Erica, Keefe Lisa J, Bowman Sarah E J

2021-Apr-01

crystal imaging, crystallization, machine learning, open-source graphical user interfaces

Internal Medicine Internal Medicine

Prediction of Disease Progression of COVID-19 Based upon Machine Learning.

In International journal of general medicine

Background : Since December 2019, COVID-19 has spread throughout the world. Clinical outcomes of COVID-19 patients vary among infected individuals. Therefore, it is vital to identify patients at high risk of disease progression.

Methods : In this retrospective, multicenter cohort study, COVID-19 patients from Huoshenshan Hospital and Taikang Tongji Hospital (Wuhan, China) were included. Clinical features showing significant differences between the severe and nonsevere groups were screened out by univariate analysis. Then, these features were used to generate classifier models to predict whether a COVID-19 case would be severe or nonsevere based on machine learning. Two test sets of data from the two hospitals were gathered to evaluate the predictive performance of the models.

Results : A total of 455 patients were included, and 21 features showing significant differences between the severe and nonsevere groups were selected for the training and validation set. The optimal subset, with eleven features in the k-nearest neighbor model, obtained the highest area under the curve (AUC) value among the four models in the validation set. D-dimer, CRP, and age were the three most important features in the optimal-feature subsets. The highest AUC value was obtained using a support vector-machine model for a test set from Huoshenshan Hospital. Software for predicting disease progression based on machine learning was developed.

Conclusion : The predictive models were successfully established based on machine learning, and achieved satisfactory predictive performance of disease progression with optimal-feature subsets.

Xu Fumin, Chen Xiao, Yin Xinru, Qiu Qiu, Xiao Jingjing, Qiao Liang, He Mi, Tang Liang, Li Xiawei, Zhang Qiao, Lv Yanling, Xiao Shili, Zhao Rong, Guo Yan, Chen Mingsheng, Chen Dongfeng, Wen Liangzhi, Wang Bin, Nian Yongjian, Liu Kaijun

2021

COVID-19, disease progression, machine-learning models

Pathology Pathology

AI-based pathology predicts origins for cancers of unknown primary.

In Nature ; h5-index 368.0

Cancer of unknown primary (CUP) origin is an enigmatic group of diagnoses in which the primary anatomical site of tumour origin cannot be determined1,2. This poses a considerable challenge, as modern therapeutics are predominantly specific to the primary tumour3. Recent research has focused on using genomics and transcriptomics to identify the origin of a tumour4-9. However, genomic testing is not always performed and lacks clinical penetration in low-resource settings. Here, to overcome these challenges, we present a deep-learning-based algorithm-Tumour Origin Assessment via Deep Learning (TOAD)-that can provide a differential diagnosis for the origin of the primary tumour using routinely acquired histology slides. We used whole-slide images of tumours with known primary origins to train a model that simultaneously identifies the tumour as primary or metastatic and predicts its site of origin. On our held-out test set of tumours with known primary origins, the model achieved a top-1 accuracy of 0.83 and a top-3 accuracy of 0.96, whereas on our external test set it achieved top-1 and top-3 accuracies of 0.80 and 0.93, respectively. We further curated a dataset of 317 cases of CUP for which a differential diagnosis was assigned. Our model predictions resulted in concordance for 61% of cases and a top-3 agreement of 82%. TOAD can be used as an assistive tool to assign a differential diagnosis to complicated cases of metastatic tumours and CUPs and could be used in conjunction with or in lieu of ancillary tests and extensive diagnostic work-ups to reduce the occurrence of CUP.

Lu Ming Y, Chen Tiffany Y, Williamson Drew F K, Zhao Melissa, Shady Maha, Lipkova Jana, Mahmood Faisal

2021-May-05

General General

Use of deep learning to develop continuous-risk models for adverse event prediction from electronic health records.

In Nature protocols

Early prediction of patient outcomes is important for targeting preventive care. This protocol describes a practical workflow for developing deep-learning risk models that can predict various clinical and operational outcomes from structured electronic health record (EHR) data. The protocol comprises five main stages: formal problem definition, data pre-processing, architecture selection, calibration and uncertainty, and generalizability evaluation. We have applied the workflow to four endpoints (acute kidney injury, mortality, length of stay and 30-day hospital readmission). The workflow can enable continuous (e.g., triggered every 6 h) and static (e.g., triggered at 24 h after admission) predictions. We also provide an open-source codebase that illustrates some key principles in EHR modeling. This protocol can be used by interdisciplinary teams with programming and clinical expertise to build deep-learning prediction models with alternate data sources and prediction tasks.

Tomašev Nenad, Harris Natalie, Baur Sebastien, Mottram Anne, Glorot Xavier, Rae Jack W, Zielinski Michal, Askham Harry, Saraiva Andre, Magliulo Valerio, Meyer Clemens, Ravuri Suman, Protsyuk Ivan, Connell Alistair, Hughes Cían O, Karthikesalingam Alan, Cornebise Julien, Montgomery Hugh, Rees Geraint, Laing Chris, Baker Clifton R, Osborne Thomas F, Reeves Ruth, Hassabis Demis, King Dominic, Suleyman Mustafa, Back Trevor, Nielson Christopher, Seneviratne Martin G, Ledsam Joseph R, Mohamed Shakir

2021-May-05

General General

Application of machine learning to large in vitro databases to identify drug-cancer cell interactions: azithromycin and KLK6 mutation status.

In Oncogene ; h5-index 102.0

Recent advances in machine learning promise to yield novel insights by interrogation of large datasets ranging from gene expression and mutation data to CRISPR knockouts and drug screens. We combined existing and new algorithms with available experimental data to identify potentially clinically relevant relationships to provide a proof of principle for the promise of machine learning in oncological drug discovery. Specifically, we screened cell line data from the Cancer Dependency Map for the effects of azithromycin, which has been shown to kill cancer cells in vitro. Our findings demonstrate a strong relationship between Kallikrein Related Peptidase 6 (KLK6) mutation status and the ability of azithromycin to kill cancer cells in vitro. While the application of azithromycin showed no meaningful average effect in KLK6 wild-type cell lines, statistically significant enhancements of cell death are seen in multiple independent KLK6-mutated cancer cell lines. These findings suggest a potentially valuable clinical strategy in patients with KLK6-mutated malignancies.

Sherman Jeff, Verstandig Grant, Rowe John W, Brumer Yisroel

2021-May-05

General General

Enhanced NH3 and H2 gas sensing with H2S gas interference using multilayer SnO2/Pt/WO3 nanofilms.

In Journal of hazardous materials

The selective detection and classification of NH3 and H2S gases with H2S gas interference based on conventional SnO2 thin film sensors is still the main problem. In this work, three layers of SnO2/Pt/WO3 nanofilms with different WO3 thicknesses (50, 80, 140, and 260 nm) were fabricated using the sputtering technique. The WO3 top layer were used as a gas filter to further improve the selectivity of sensors. The effect of WO3 thickness on the (NH3, H2, and H2S) gas-sensing properties of the sensors was investigated. At the optimal WO3 thickness of 140 nm, the gas responses of SnO2/Pt/WO3 sensors toward NH3 and H2 gases were slightly lower than those of Pt/SnO2 sensor film, and the gas response of SnO2/Pt/WO3 sensor films to H2S gas was almost negligible. The calcification of NH3 and H2 gases was effectively conducted by machine learning algorithms. These evidences manifested that SnO2/Pt/WO3 sensor films are suitable for the actual NH3 detection of NH3 and H2S gases.

Van Toan Nguyen, Hung Chu Manh, Hoa Nguyen Duc, Van Duy Nguyen, Thi Thanh Le Dang, Thi Thu Hoa Nguyen, Viet Nguyen Ngoc, Phuoc Phan Hong, Van Hieu Nguyen

2021-Jun-15

Gas filter membrane, H(2)S, NH(3), Pt/SnO(2) nanofilm, WO(3) layer

Internal Medicine Internal Medicine

Prediction of Disease Progression of COVID-19 Based upon Machine Learning.

In International journal of general medicine

Background : Since December 2019, COVID-19 has spread throughout the world. Clinical outcomes of COVID-19 patients vary among infected individuals. Therefore, it is vital to identify patients at high risk of disease progression.

Methods : In this retrospective, multicenter cohort study, COVID-19 patients from Huoshenshan Hospital and Taikang Tongji Hospital (Wuhan, China) were included. Clinical features showing significant differences between the severe and nonsevere groups were screened out by univariate analysis. Then, these features were used to generate classifier models to predict whether a COVID-19 case would be severe or nonsevere based on machine learning. Two test sets of data from the two hospitals were gathered to evaluate the predictive performance of the models.

Results : A total of 455 patients were included, and 21 features showing significant differences between the severe and nonsevere groups were selected for the training and validation set. The optimal subset, with eleven features in the k-nearest neighbor model, obtained the highest area under the curve (AUC) value among the four models in the validation set. D-dimer, CRP, and age were the three most important features in the optimal-feature subsets. The highest AUC value was obtained using a support vector-machine model for a test set from Huoshenshan Hospital. Software for predicting disease progression based on machine learning was developed.

Conclusion : The predictive models were successfully established based on machine learning, and achieved satisfactory predictive performance of disease progression with optimal-feature subsets.

Xu Fumin, Chen Xiao, Yin Xinru, Qiu Qiu, Xiao Jingjing, Qiao Liang, He Mi, Tang Liang, Li Xiawei, Zhang Qiao, Lv Yanling, Xiao Shili, Zhao Rong, Guo Yan, Chen Mingsheng, Chen Dongfeng, Wen Liangzhi, Wang Bin, Nian Yongjian, Liu Kaijun

2021

COVID-19, disease progression, machine-learning models

oncology Oncology

Identification of diagnostic markers and lipid dysregulation in oesophageal squamous cell carcinoma through lipidomic analysis and machine learning.

In British journal of cancer ; h5-index 89.0

BACKGROUND : Oesophageal cancer (EC) ranks high in both morbidity and mortality. A non-invasive and high-sensitivity diagnostic approach is necessary to improve the prognosis of EC patients.

METHODS : A total of 525 serum samples were subjected to lipidomic analysis. We combined serum lipidomics and machine-learning algorithms to select important metabolite features for the detection of oesophageal squamous cell carcinoma (ESCC), the major subtype of EC in developing countries. A diagnostic model using a panel of selected features was developed and evaluated. Integrative analyses of tissue transcriptome and serum lipidome were conducted to reveal the underlying mechanism of lipid dysregulation.

RESULTS : Our optimised diagnostic model with a panel of 12 lipid biomarkers together with age and gender reaches a sensitivity of 90.7%, 91.3% and 90.7% and an area under receiver-operating characteristic curve of 0.958, 0.966 and 0.818 in detecting ESCC for the training cohort, validation cohort and independent validation cohort, respectively. Integrative analysis revealed matched variation trend of genes encoding key enzymes in lipid metabolism.

CONCLUSIONS : We have identified a panel of 12 lipid biomarkers for diagnostic modelling and potential mechanisms of lipid dysregulation in the serum of ESCC patients. This is a reliable, rapid and non-invasive tumour-diagnostic approach for clinical application.

Yuan Yuyao, Zhao Zitong, Xue Liyan, Wang Guangxi, Song Huajie, Pang Ruifang, Zhou Juntuo, Luo Jianyuan, Song Yongmei, Yin Yuxin

2021-May-05

General General

A deep explainable artificial intelligent framework for neurological disorders discrimination.

In Scientific reports ; h5-index 158.0

Pathological hand tremor (PHT) is a common symptom of Parkinson's disease (PD) and essential tremor (ET), which affects manual targeting, motor coordination, and movement kinetics. Effective treatment and management of the symptoms relies on the correct and in-time diagnosis of the affected individuals, where the characteristics of PHT serve as an imperative metric for this purpose. Due to the overlapping features of the corresponding symptoms, however, a high level of expertise and specialized diagnostic methodologies are required to correctly distinguish PD from ET. In this work, we propose the data-driven [Formula: see text] model, which processes the kinematics of the hand in the affected individuals and classifies the patients into PD or ET. [Formula: see text] is trained over 90 hours of hand motion signals consisting of 250 tremor assessments from 81 patients, recorded at the London Movement Disorders Centre, ON, Canada. The [Formula: see text] outperforms its state-of-the-art counterparts achieving exceptional differential diagnosis accuracy of [Formula: see text]. In addition, using the explainability and interpretability measures for machine learning models, clinically viable and statistically significant insights on how the data-driven model discriminates between the two groups of patients are achieved.

Shahtalebi Soroosh, Atashzar S Farokh, Patel Rajni V, Jog Mandar S, Mohammadi Arash

2021-May-05

General General

Deep convolution stack for waveform in underwater acoustic target recognition.

In Scientific reports ; h5-index 158.0

In underwater acoustic target recognition, deep learning methods have been proved to be effective on recognizing original signal waveform. Previous methods often utilize large convolutional kernels to extract features at the beginning of neural networks. It leads to a lack of depth and structural imbalance of networks. The power of nonlinear transformation brought by deep network has not been fully utilized. Deep convolution stack is a kind of network frame with flexible and balanced structure and it has not been explored well in underwater acoustic target recognition, even though such frame has been proven to be effective in other deep learning fields. In this paper, a multiscale residual unit (MSRU) is proposed to construct deep convolution stack network. Based on MSRU, a multiscale residual deep neural network (MSRDN) is presented to classify underwater acoustic target. Dataset acquired in a real-world scenario is used to verify the proposed unit and model. By adding MSRU into Generative Adversarial Networks, the validity of MSRU is proved. Finally, MSRDN achieves the best recognition accuracy of 83.15%, improved by 6.99% from the structure related networks which take the original signal waveform as input and 4.48% from the networks which take the time-frequency representation as input.

Tian Shengzhao, Chen Duanbing, Wang Hang, Liu Jingfa

2021-May-05

General General

Selection of nitrogen responsive root architectural traits in spinach using machine learning and genetic correlations.

In Scientific reports ; h5-index 158.0

The efficient acquisition and transport of nutrients by plants largely depend on the root architecture. Due to the absence of complex microbial network interactions and soil heterogeneity in a restricted soilless medium, the architecture of roots is a function of genetics defined by the soilless matrix and exogenously supplied nutrients such as nitrogen (N). The knowledge of root trait combinations that offer the optimal nitrogen use efficiency (NUE) is far from being conclusive. The objective of this study was to define the root trait(s) that best predicts and correlates with vegetative biomass under differed N treatments. We used eight image-derived root architectural traits of 202 diverse spinach lines grown in two N concentrations (high N, HN, and low N, LN) in randomized complete blocks design. Supervised random forest (RF) machine learning augmented by ranger hyperparameter grid search was used to predict the variable importance of the root traits. We also determined the broad-sense heritability (H) and genetic (rg) and phenotypic (rp) correlations between root traits and the vegetative biomass (shoot weight, SWt). Each root trait was assigned a predicted importance rank based on the trait's contribution to the cumulative reduction in the mean square error (MSE) in the RF tree regression models for SWt. The root traits were further prioritized for potential selection based on the rg and SWt correlated response (CR). The predicted importance of the eight root traits showed that the number of root tips (Tips) and root length (RLength) under HN and crossings (Xsings) and root average diameter (RAvdiam) under LN were the most relevant. SWt had a highly antagonistic rg (- 0.83) to RAvdiam, but a high predicted indirect selection efficiency (- 112.8%) with RAvdiam under LN; RAvdiam showed no significant rg or rp to SWt under HN. In limited N availability, we suggest that selecting against larger RAvdiam as a secondary trait might improve biomass and, hence, NUE with no apparent yield penalty under HN.

Awika Henry O, Mishra Amit K, Gill Haramrit, DiPiazza James, Avila Carlos A, Joshi Vijay

2021-May-05

General General

Deep hierarchical embedding for simultaneous modeling of GPCR proteins in a unified metric space.

In Scientific reports ; h5-index 158.0

GPCR proteins belong to diverse families of proteins that are defined at multiple hierarchical levels. Inspecting relationships between GPCR proteins on the hierarchical structure is important, since characteristics of the protein can be inferred from proteins in similar hierarchical information. However, modeling of GPCR families has been performed separately for each of the family, subfamily, and sub-subfamily level. Relationships between GPCR proteins are ignored in these approaches as they process the information in the proteins with several disconnected models. In this study, we propose DeepHier, a deep learning model to simultaneously learn representations of GPCR family hierarchy from the protein sequences with a unified single model. Novel loss term based on metric learning is introduced to incorporate hierarchical relations between proteins. We tested our approach using a public GPCR sequence dataset. Metric distances in the deep feature space corresponded to the hierarchical family relation between GPCR proteins. Furthermore, we demonstrated that further downstream tasks, like phylogenetic reconstruction and motif discovery, are feasible in the constructed embedding space. These results show that hierarchical relations between sequences were successfully captured in both of technical and biological aspects.

Lee Taeheon, Lee Sangseon, Kang Minji, Kim Sun

2021-May-05

General General

Regression plane concept for analysing continuous cellular processes with machine learning.

In Nature communications ; h5-index 260.0

Biological processes are inherently continuous, and the chance of phenotypic discovery is significantly restricted by discretising them. Using multi-parametric active regression we introduce the Regression Plane (RP), a user-friendly discovery tool enabling class-free phenotypic supervised machine learning, to describe and explore biological data in a continuous manner. First, we compare traditional classification with regression in a simulated experimental setup. Second, we use our framework to identify genes involved in regulating triglyceride levels in human cells. Subsequently, we analyse a time-lapse dataset on mitosis to demonstrate that the proposed methodology is capable of modelling complex processes at infinite resolution. Finally, we show that hemocyte differentiation in Drosophila melanogaster has continuous characteristics.

Szkalisity Abel, Piccinini Filippo, Beleon Attila, Balassa Tamas, Varga Istvan Gergely, Migh Ede, Molnar Csaba, Paavolainen Lassi, Timonen Sanna, Banerjee Indranil, Ikonen Elina, Yamauchi Yohei, Ando Istvan, Peltonen Jaakko, Pietiäinen Vilja, Honti Viktor, Horvath Peter

2021-05-05

Surgery Surgery

Digital synthesis of histological stains using micro-structured and multiplexed virtual staining of label-free tissue.

In Light, science & applications

Histological staining is a vital step in diagnosing various diseases and has been used for more than a century to provide contrast in tissue sections, rendering the tissue constituents visible for microscopic analysis by medical experts. However, this process is time consuming, labour intensive, expensive and destructive to the specimen. Recently, the ability to virtually stain unlabelled tissue sections, entirely avoiding the histochemical staining step, has been demonstrated using tissue-stain-specific deep neural networks. Here, we present a new deep-learning-based framework that generates virtually stained images using label-free tissue images, in which different stains are merged following a micro-structure map defined by the user. This approach uses a single deep neural network that receives two different sources of information as its input: (1) autofluorescence images of the label-free tissue sample and (2) a "digital staining matrix", which represents the desired microscopic map of the different stains to be virtually generated in the same tissue section. This digital staining matrix is also used to virtually blend existing stains, digitally synthesizing new histological stains. We trained and blindly tested this virtual-staining network using unlabelled kidney tissue sections to generate micro-structured combinations of haematoxylin and eosin (H&E), Jones' silver stain, and Masson's trichrome stain. Using a single network, this approach multiplexes the virtual staining of label-free tissue images with multiple types of stains and paves the way for synthesizing new digital histological stains that can be created in the same tissue cross section, which is currently not feasible with standard histochemical staining methods.

Zhang Yijie, de Haan Kevin, Rivenson Yair, Li Jingxi, Delis Apostolos, Ozcan Aydogan

2020-May-06

Ophthalmology Ophthalmology

Literature Commentary.

In Journal of neuro-ophthalmology : the official journal of the North American Neuro-Ophthalmology Society

In this issue of Journal of Neuro-Ophthalmology, M. Tariq Bhatti, MD, and Mark L. Moster, MD will discuss the following 6 articles: Powell G, Derry-Sumner H, Rajenderkumar D, Rushton SK, Sumner P. Persistent postural perceptual dizziness is on a spectrum in the general population. Neurology. 2020;94:e1929-e1938.Mackie SL, Dejaco C, Appenzeller S, Camellino D, Duftner C, Gonzalez-Chiappe S, Mahr A, Mukhtyar C, Reynolds G, de Souza AWS, Brouwer E, Bukhari M, Buttgereit F, Byrne D, Cid MC, Cimmino M, Direskeneli H, Gilbert K, Kermani TA, Khan A, Lanyon P, Luqmani R, Mallen C, Mason JC, Matteson EL, Merkel PA, Mollan S, Neill L, Sullivan E, Sandovici M, Schmidt WA, Watts R, Whitlock M, Yacyshyn E, Ytterberg S, Dasgupta B. British Society for Rheumatology guideline on diagnosis and treatment of giant cell arteritis: executive summary. Rheumatology (Oxford). 2020;59:487-494.Yang HK, Kim YJ, Sung JY, Kim DH, Kim KG, Hwang JM. Efficacy for differentiating nonglaucomatous vs glaucomatous optic neuropathy using deep learning systems. Am J Ophthalmol. [published online ahead of print April 2, 2020] doi:10.1016/j.ajo.2020.03.035.Milea D, Najjar RP, Zhubo J, Ting D, Vasseneix C, Xu X, Fard MA, Fonseca P, Vanikieti K, Lagrèze WA, Morgia CL, Cheung CY, Hamann S, Chiquet C, Sanda N, Yang H, Mejico LJ, Rougier MB, Kho R, Tran THC, Singhal S, Gohier P, Clermont-Vignal C, Cheng CY, Jonas JB, Yu-Wai-Man P, Fraser CL, Chen JJ, Ambika S, Miller NR, Liu Y, Newman NJ, Wong TY, Biousse V, the BONSAI Group. Artificial intelligence to detect papilledema from ocular fundus photographs. N Engl J Med. 2020;382:1687-1695.Ghanem KG, Ram S, Rice PA. The modern epidemic of syphilis. N Engl J Med. 2020;382:845-854.Woolen SA, Shankar PR, Gagnier JJ, MacEachern MP, Singer L, Davenport MS. Risk of nephrogenic systemic fibrosis in patients with stage 4 or 5 chronic kidney disease receiving a group II gadolinium-based contrast agent: a systematic review and meta-analysis. JAMA Intern Med. 2019;180:223-230.

**

2020-Sep-01

oncology Oncology

Tumor-associated Stromal Cellular Density as a Predictor of Recurrence and Mortality in Breast Cancer: Results from Ethnically-diverse Study Populations.

In Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology

BACKGROUND : Tumor-associated stroma is comprised of fibroblasts, tumor infiltrating lymphocytes (TILs), macrophages, endothelial, and other cells, that interactively influence tumor progression through inflammation and wound repair. Although gene expression signatures reflecting wound repair predict breast cancer survival, it is unclear whether combined density of tumor-associated stromal cells, a morphological proxy for inflammation and wound repair signatures on routine hematoxylin and eosin (H&E)-stained sections, is of prognostic relevance.

METHODS : By applying machine learning to digitized H&E-stained sections for 2,084 breast cancer patients from China (n=596; 24-55years), Poland (n=810; 31-75years), and the United States (n=678; 55-78years), we characterized tumor-associated stromal cellular density (SCD) as the percentage of tumor-stroma that is occupied by nucleated cells. Hazard ratios (HR) and 95% confidence intervals (CI) for associations between SCD and clinical outcomes (recurrence (China); mortality (Poland and United States)) were estimated using Cox proportional hazard regression, adjusted for clinical variables.

RESULTS : SCD was independently predictive of poor clinical outcomes in hormone receptor-positive (luminal) tumors from China (multivariable HR(95%CI)fourth(Q4) vs first(Q1) quartile=1.86(1.06-3.26);Ptrend=0.03), Poland (HR(95%CI)Q4 vs Q1=1.80(1.12-2.89);Ptrend=0.01), and United States (HR(95%CI)Q4 vs Q1=2.42(1.33-4.42);Ptrend=0.002). In general, SCD provided more prognostic information than most classical clinicopathologic factors, including grade, size, PR, HER2, IHC4, and TILs, predicting clinical outcomes irrespective of menopausal or lymph nodal status. SCD was not predictive of outcomes in hormone receptor-negative tumors.

CONCLUSIONS : Our findings support the independent prognostic value of tumor-associated SCD among ethnically-diverse luminal breast cancer patients.

IMPACT : Assessment of tumor-associated SCD on standard H&E could help refine prognostic assessment and therapeutic decision-making in luminal breast cancer.

Abubakar Mustapha, Zhang Jing, Ahearn Thomas U, Koka Hela, Guo Changyuan, Lawrence Scott M, Mutreja Karun, Figueroa Jonine D, Ying Jianming, Lissowska Jolanta, Lyu Ning, Garcia-Closas Montserrat, Yang Xiaohong R

2021-May-05

Pathology Pathology

Evaluation of an open-source machine-learning tool to quantify bone marrow plasma cells.

In Journal of clinical pathology

AIMS : The objective of this study was to develop and validate an open-source digital pathology tool, QuPath, to automatically quantify CD138-positive bone marrow plasma cells (BMPCs).

METHODS : We analysed CD138-scanned slides in QuPath. In the initial training phase, manual positive and negative cell counts were performed in representative areas of 10 bone marrow biopsies. Values from the manual counts were used to fine-tune parameters to detect BMPCs, using the positive cell detection and neural network (NN) classifier functions. In the testing phase, whole-slide images in an additional 40 cases were analysed. Output from the NN classifier was compared with two pathologist's estimates of BMPC percentage.

RESULTS : The training set included manual counts ranging from 2403 to 17 287 cells per slide, with a median BMPC percentage of 13% (range: 3.1%-80.7%). In the testing phase, the quantification of plasma cells by image analysis correlated well with manual counting, particularly when restricted to BMPC percentages of <30% (Pearson's r=0.96, p<0.001). Concordance between the NN classifier and the pathologist whole-slide estimates was similarly good, with an intraclass correlation of 0.83 and a weighted kappa for the NN classifier of 0.80 with the first rater and 0.90 with the second rater. This was similar to the weighted kappa between the two human raters (0.81).

CONCLUSIONS : This represents a validated digital pathology tool to assist in automatically and reliably counting BMPC percentage on CD138-stained slides with an acceptable error rate.

Baranova Katherina, Tran Christopher, Plantinga Paul, Sangle Nikhil

2021-May-05

bone marrow neoplasms, computer-assisted, image processing, multiple myeloma, pathology, surgical

General General

Application of Novel Machine Learning Techniques for Predicting the Surface Chloride Concentration in Concrete Containing Waste Material.

In Materials (Basel, Switzerland)

Structures located on the coast are subjected to the long-term influence of chloride ions, which cause the corrosion of steel reinforcements in concrete elements. This corrosion severely affects the performance of the elements and may shorten the lifespan of an entire structure. Even though experimental activities in laboratories might be a solution, they may also be problematic due to time and costs. Thus, the application of individual machine learning (ML) techniques has been investigated to predict surface chloride concentrations (Cc) in marine structures. For this purpose, the values of Cc in tidal, splash, and submerged zones were collected from an extensive literature survey and incorporated into the article. Gene expression programming (GEP), the decision tree (DT), and an artificial neural network (ANN) were used to predict the surface chloride concentrations, and the most accurate algorithm was then selected. The GEP model was the most accurate when compared to ANN and DT, which was confirmed by the high accuracy level of the K-fold cross-validation and linear correlation coefficient (R2), mean absolute error (MAE), mean square error (MSE), and root mean square error (RMSE) parameters. As is shown in the article, the proposed method is an effective and accurate way to predict the surface chloride concentration without the inconveniences of laboratory tests.

Ahmad Ayaz, Farooq Furqan, Ostrowski Krzysztof Adam, Śliwa-Wieczorek Klaudia, Czarnecki Slawomir

2021-Apr-29

aggressive ions environment, artificial neural networks, concrete, gene expression programming, individual algorithm, surface chloride concentration

Surgery Surgery

Speech Analysis Using Artificial Intelligence as a Peri-Operative Evaluation: A Case Report of a Patient with Temporal Lobe Epilepsy Secondary to Tuberous Sclerosis Complex Who Underwent Epilepsy Surgery.

In Brain sciences

BACKGROUND : Improved conversational fluency is sometimes identified postoperatively in patients with epilepsy, but improvements can be difficult to assess using tests such as the intelligence quotient (IQ) test. Evaluation of pre- and postoperative differences might be considered subjective at present because of the lack of objective criteria. Artificial intelligence (AI) could possibly be used to make the evaluations more objective. The aim of this case report is thus to analyze the speech of a young female patient with epilepsy before and after surgery.

METHOD : The speech of a nine-year-old girl with epilepsy secondary to tuberous sclerosis complex is recorded during interviews one month before and two months after surgery. The recorded speech is then manually transcribed and annotated, and subsequently automatically analyzed using AI software. IQ testing is also conducted on both occasions. The patient remains seizure-free for at least 13 months postoperatively.

RESULTS : There are decreases in total interview time and subjective case markers per second, whereas there are increases in morphemes and objective case markers per second. Postoperatively, IQ scores improve, except for the Perceptual Reasoning Index.

CONCLUSIONS : AI analysis is able to identify differences in speech before and after epilepsy surgery upon an epilepsy patient with tuberous sclerosis complex.

Niimi Keiko, Fujimoto Ayataka, Kano Yoshinobu, Otsuki Yoshiro, Enoki Hideo, Okanishi Tohru

2021-Apr-29

artificial intelligence, epilepsy, intelligence quotient, speech analysis, surgery

oncology Oncology

Endoscopist Diagnostic Accuracy in Detecting Upper-GI Neoplasia in the Framework of Artificial Intelligence Studies.

In Endoscopy ; h5-index 58.0

AIMS Estimates on miss-rates for upper gastrointestinal neoplasia (UGIN) rely on registry data or old studies. Quality assurance programs for upper-GI endoscopy are not fully established due to the lack of the infrastructure to measure endoscopists' competence. We aimed at assessing endoscopists' accuracy for the recognition of UGIN exploiting the framework of Artificial Intelligence (AI) validation studies. METHODS Literature search among databases (PubMed/MEDLINE, EMBASE, Scopus) up to August 2020 was performed to identify articles evaluating the accuracy of individual endoscopists for the recognition of UGIN within studies validating AI against a histologically-verified expert-annotated ground-truth. Main outcomes were endoscopists' pooled sensitivity, specificity, positive and negative predictive values (PPV/NPV), Area Under the Curve (AUCs) for all-UGIN, for esophageal squamous-cell neoplasia (ESCN), Barrett-related neoplasia (BERN), and gastric adenocarcinoma (GAC). RESULTS Seven studies with 122 endoscopists were included (2 ESCN, 3 BERN, 1 GAC, 1 UGIN overall). Pooled endoscopists sensitivity and specificity for UGIN was 82% (CI 80-84%) and 79% (CI 76-81%), respectively. Endoscopists' accuracy was higher for GAC detection (AUC 0.95, CI 0.93-0.98) than ESCN detection (AUC 0.90, CI 0.88-0.92) and BERN detection (AUC 0.86, CI0.84-0.88). Sensitivity was higher for Eastern vs. Western endoscopists (87%, CI 84-89% vs. 75%, CI 72-78%), and for experts vs. non-experts endoscopists (85%, CI 83-87% vs. 71%, CI 67-75%). CONCLUSION We show suboptimal endoscopists' accuracy for the recognition of UGIN even within a framework that included higher prevalence and disease awareness. Future AI validation studies represent a framework to assess endoscopist competence.

Frazzoni Leonardo, Arribas Julia, Antonelli Giulio, Libânio Diogo, Ebigbo Alanna, van der Sommen Fons, De Groof Albert Jeroen, Fukuda Hiromu, Ohmori Masayasu, Ishihara Ryu, Wu Lianlian, Yu Honggang, Repici Alessandro, Bergman Jacques J G H M, Sharma Prateek, Messmann Helmut, Hassan Cesare, Fuccio Lorenzo, Dinis-Ribeiro Mário

2021-May-05

General General

SSI-DDI: substructure-substructure interactions for drug-drug interaction prediction.

In Briefings in bioinformatics

A major concern with co-administration of different drugs is the high risk of interference between their mechanisms of action, known as adverse drug-drug interactions (DDIs), which can cause serious injuries to the organism. Although several computational methods have been proposed for identifying potential adverse DDIs, there is still room for improvement. Existing methods are not explicitly based on the knowledge that DDIs are fundamentally caused by chemical substructure interactions instead of whole drugs' chemical structures. Furthermore, most of existing methods rely on manually engineered molecular representation, which is limited by the domain expert's knowledge.We propose substructure-substructure interaction-drug-drug interaction (SSI-DDI), a deep learning framework, which operates directly on the raw molecular graph representations of drugs for richer feature extraction; and, most importantly, breaks the DDI prediction task between two drugs down to identifying pairwise interactions between their respective substructures. SSI-DDI is evaluated on real-world data and improves DDI prediction performance compared to state-of-the-art methods. Source code is freely available at https://github.com/kanz76/SSI-DDI.

Nyamabo Arnold K, Yu Hui, Shi Jian-Yu

2021-May-05

co-attention, drug–drug interactions, molecular graph, multi-type interactions, substructure interactions

General General

Two degree-of-freedom robotic eye: design, modeling, and learning-based control in foveation and smooth pursuit.

In Bioinspiration & biomimetics

With increasing ocular motility disorders affecting human eye movement, the need to understand the biomechanics of the human eye rises constantly. A robotic eye system that physically mimics the human eye can serve as a useful tool for biomedical researchers to obtain an intuitive understanding of the functions and defects of the extraocular muscles and the eye. This paper presents the design, modeling, and control of a two degree-of-freedom (2-DOF) robotic eye, driven by artificial muscles, in particular, made of super-coiled polymers (SCPs). Considering the highly nonlinear dynamics of the robotic eye system, this paper applies deep deterministic policy gradient (DDPG), a machine learning algorithm to solve the control design problem in foveation and smooth pursuit of the robotic eye. To the best of our knowledge, this paper presents the first modeling effort to establish the dynamics of a robotic eye driven by SCP actuators, as well as the first control design effort for robotic eyes using a DDPG-based control strategy. A linear quadratic regulator (LQR)-type reward function is proposed to achieve a balance between system performances (convergence speed and tracking accuracy) and control efforts. Simulation results are presented to demonstrate the effectiveness of the proposed control strategy for the 2-DOF robotic eye.

Rajendran Sunil Kumar, Wei Qi, Zhang Feitian

2021-May-05

Biologically-Inspired Robots, Biomimetics, Dynamics, Machine Learning for Robot Control, Robotic Eye

General General

Is Human Walking a Network Medicine Problem? An Analysis Using Symbolic Regression Models with Genetic Programming.

In Computer methods and programs in biomedicine

BACKGROUND AND OBJECTIVE : Human walking is typically assessed using a sensor placed on the lower back or the hip. Such analyses often ignore that the arms, legs, and body trunk movements all have significant roles during walking; in other words, these body nodes with accelerometers form a body sensor network (BSN). BSN refers to a network of wearable sensors or devices on the human body that collects physiological signals. Our study proposes that human locomotion could be considered as a network of well-connected nodes.

METHODS : While hypothesizing that accelerometer data can model this BSN, we collected accelerometer signals from six body areas from ten healthy participants performing a cognitive task. Machine learning based on genetic programming was used to produce a collection of non-linear symbolic models of human locomotion.

RESULTS : With implications in precision medicine, our primary finding was that our BSN models fit the data from the lower back's accelerometer and describe subject-specific data the best compared to all other models. Across subjects, models were less effective due to the diversity of human sizes.

CONCLUSIONS : A BSN relationship between all six body nodes has been shown to describe the subject-specific data, which indicates that the network-medicine relationship between these nodes is essential in adequately describing human walking. Our gait analyses can be used for several clinical applications such as medical diagnostics as well as creating a baseline for healthy walking with and without a cognitive load.

Dasgupta Pritika, Hughes James Alexander, Daley Mark, Sejdić Ervin

2021-Apr-10

acceleration gait measures, genetic programming, mathematical model, symbolic regression, walking, wearables

General General

Predicting youth at high risk of aging out of foster care using machine learning methods.

In Child abuse & neglect

BACKGROUND : Youth who exit the nation's foster care system without permanency are at high risk of experiencing difficulties during the transition to adulthood.

OBJECTIVE : To present an illustrative test of whether an algorithmic decision aid could be used to identify youth at risk of existing foster care without permanency.

METHODS : For youth placed in foster care between ages 12 and 14, we assessed the risk of exiting care without permanency by age 18 based on their child welfare service involvement history. To develop predictive risk models, 28 years (1991-2018) of child welfare service records from California were used. Performances were evaluated using F1, AUC, and precision and recall scores at k %. Algorithmic racial bias and fairness was also examined.

RESULTS : The gradient boosting decision tree and random forest showed the best performance (F1 score = .54-.55, precision score = .62, recall score = .49). Among the top 30 % of youth the model identified as high risk, half of all youth who exited care without permanency were accurately identified four to six years prior to their exit, with a 39 % error rate. Although racial disparities between Black and White youth were observed in imbalanced error rates, calibration and predictive parity were satisfied.

CONCLUSIONS : Our study illustrates the manner in which potential applications of predictive analytics, including those designed to achieve universal goals of permanency through more targeted allocations of resources, can be tested. It also assesses the model using metrics of fairness.

Ahn Eunhye, Gil Yolanda, Putnam-Hornstein Emily

2021-May-02

Algorithmic fairness, Machine learning, Permanency, Predictive modeling, Transitional support programs, Youth in foster care

General General

Random Forest Modelling of Milk Yield of Dairy Cows under Heat Stress Conditions.

In Animals : an open access journal from MDPI

Precision Livestock Farming (PLF) relies on several technological approaches to acquire, in the most efficient way, precise and real-time data concerning production and welfare of individual animals. In this regard, in the dairy sector, PLF devices are being increasingly adopted, automatic milking systems (AMSs) are becoming increasingly widespread, and monitoring systems for animals and environmental conditions are becoming common tools in herd management. As a consequence, a great amount of daily recorded data concerning individual animals are available for the farmers and they could be used effectively for the calibration of numerical models to be used for the prediction of future animal production trends. On the other hand, the machine learning approaches in PLF are nowadays considered an extremely promising solution in the research field of livestock farms and the application of these techniques in the dairy cattle farming would increase sustainability and efficiency of the sector. The study aims to define, train, and test a model developed through machine learning techniques, adopting a Random Forest algorithm, having the main goal to assess the trend in daily milk yield of a single cow in relation to environmental conditions. The model has been calibrated and tested on the data collected on 91 lactating cows of a dairy farm, located in northern Italy, and equipped with an AMS and thermo-hygrometric sensors during the years 2016-2017. In the statistical model, having seven predictor features, the daily milk yield is evaluated as a function of the position of the day in the lactation curve and the indoor barn conditions expressed in terms of daily average of the temperature-humidity index (THI) in the same day and its value in each of the five previous days. In this way, extreme hot conditions inducing heat stress effects can be considered in the yield predictions by the model. The average relative prediction error of the milk yield of each cow is about 18% of daily production, and only 2% of the total milk production.

Bovo Marco, Agrusti Miki, Benni Stefano, Torreggiani Daniele, Tassinari Patrizia

2021-Apr-30

heat stress, livestock sustainability, machine learning, precision livestock farming, random forest

General General

Convolutional neural networks for the diagnosis and prognosis of the coronavirus disease pandemic.

In Visual computing for industry, biomedicine, and art

A neural network is one of the current trends in deep learning, which is increasingly gaining attention owing to its contribution in transforming the different facets of human life. It also paves a way to approach the current crisis caused by the coronavirus disease (COVID-19) from all scientific directions. Convolutional neural network (CNN), a type of neural network, is extensively applied in the medical field, and is particularly useful in the current COVID-19 pandemic. In this article, we present the application of CNNs for the diagnosis and prognosis of COVID-19 using X-ray and computed tomography (CT) images of COVID-19 patients. The CNN models discussed in this review were mainly developed for the detection, classification, and segmentation of COVID-19 images. The base models used for detection and classification were AlexNet, Visual Geometry Group Network with 16 layers, residual network, DensNet, GoogLeNet, MobileNet, Inception, and extreme Inception. U-Net and voxel-based broad learning network were used for segmentation. Even with limited datasets, these methods proved to be beneficial for efficiently identifying the occurrence of COVID-19. To further validate these observations, we conducted an experimental study using a simple CNN framework for the binary classification of COVID-19 CT images. We achieved an accuracy of 93% with an F1-score of 0.93. Thus, with the availability of improved medical image datasets, it is evident that CNNs are very useful for the efficient diagnosis and prognosis of COVID-19.

Kugunavar Sneha, Prabhakar C J

2021-May-05

COVID-19, Convolutional neural network, Deep learning, Medical image analysis, Neural network

Cardiology Cardiology

Digital biomarkers and algorithms for detection of atrial fibrillation using surface electrocardiograms: A systematic review.

In Computers in biology and medicine

AIMS : Automated detection of atrial fibrillation (AF) in continuous rhythm registrations is essential in order to prevent complications and optimize treatment of AF. Many algorithms have been developed to detect AF in surface electrocardiograms (ECGs) during the past few years. The aim of this systematic review is to gain more insight into these available classification methods by discussing previously used digital biomarkers and algorithms and make recommendations for future research.

METHODS : On the 14th of September 2020, the PubMed database was searched for articles focusing on algorithms for AF detection in ECGs using the MeSH terms Atrial Fibrillation, Electrocardiography and Algorithms. Articles which solely focused on differentiation of types of rhythm disorders or prediction of AF termination were excluded.

RESULTS : The search resulted in 451 articles, of which 130 remained after full-text screening. Not only did the amount of research on methods for AF detection increase over the past years, but a trend towards more complex classification methods is observed. Furthermore, three different types of features can be distinguished: atrial features, ventricular features, and signal features. Although AF is an atrial disease, only 22% of the described methods use atrial features.

CONCLUSION : More and more studies focus on improving accuracy of classification methods for AF in ECGs. As a result, algorithms become increasingly complex and less well interpretable. Only a few studies focus on detecting atrial activity in the ECG. Developing innovative methods focusing on detection of atrial activity might provide accurate classifiers without compromising on transparency.

Wesselius Fons J, van Schie Mathijs S, De Groot Natasja M S, Hendriks Richard C

2021-Apr-15

Algorithms, Atrial fibrillation, Classification, ECG signal Processing, Machine learning, Telemetry

General General

Machine learning for reparameterization of four-site water models: TIP4P-BG and TIP4P-BGT.

In Physical chemistry chemical physics : PCCP

Parameterizing an effective water model is a challenging issue because of the difficulty in maintaining a comprehensive balance among the diverse physical properties of water with a limited number of parameters. The advancement in machine learning provides a promising path to search for a reliable set of parameters. Based on the TIP4P water model, hence, about 6000 molecular dynamics (MD) simulations for pure water at 1 atm and in the range of 273-373 K are conducted here as the training data. The back-propagation (BP) neural network is then utilized to construct an efficient mapping between the model parameters and four crucial physical properties of water, including the density, vaporization enthalpy, self-diffusion coefficient and viscosity. Without additional time-consuming MD simulations, this mapping operation could result in sufficient and accurate data for high-population genetic algorithm (GA) to optimize the model parameters as much as possible. Based on the proposed parameterizing strategy, TIP4P-BG (a conventional four-site water model) and TIP4P-BGT (an advanced model with temperature-dependent parameters) are established. Both the water models exhibit excellent performance with a reasonable balance among the four crucial physical properties. The relevant mean absolute percentage errors are 3.53% and 3.08%, respectively. Further calculations on the temperature of maximum density, isothermal compressibility, thermal expansion coefficient, radial distribution function and surface tension are also performed and the resulting values are in good agreement with the experimental values. Through this water modeling example, the potential of the proposed data-driven machine learning procedure has been demonstrated for parameterizing a MD-based material model.

Ye Hong-Fei, Wang Jian, Zheng Yong-Gang, Zhang Hong-Wu, Chen Zhen

2021-May-07

General General

A novel diagnostic and prognostic approach for unresponsive patients with anthroponotic cutaneous leishmaniasis using artificial neural networks.

In PloS one ; h5-index 176.0

Cutaneous leishmaniasis (CL) imposes a major health burden throughout the tropical and subtropical regions of the globe. Unresponsive cases are common phenomena occurred upon exposure to the standard drugs. Therefore, rapid detection, prognosis and classification of the disease are crucial for selecting the proper treatment modality. Using machine learning (ML) techniques, this study aimed to detect unresponsive cases of ACL, caused by Leishmania tropica, which will consequently be used for a more effective treatment modality. This study was conducted as a case-control setting. Patients were selected in a major ACL focus from both unresponsive and responsive cases. Nine unique and relevant features of patients with ACL were selected. To categorize the patients, different classifier models such as k-nearest neighbors (KNN), support vector machines (SVM), multilayer perceptron (MLP), learning vector quantization (LVQ) and multipass LVQ were applied and compared for this supervised learning task. Comparison of the receiver operating characteristic graphs (ROC) and confusion plots for the above models represented that MLP was a fairly accurate prediction model to solve this problem. The overall accuracy in terms of sensitivity, specificity and area under ROC curve (AUC) of MLP classifier were 87.8%, 90.3%, 86% and 0.88%, respectively. Moreover, the duration of the skin lesion was the most influential feature in MLP classifier, while gender was the least. The present investigation demonstrated that MLP model could be utilized for rapid detection, accurate prognosis and effective treatment of unresponsive patients with ACL. The results showed that the major feature affecting the responsiveness to treatments is the duration of the lesion. This novel approach is unique and can be beneficial in developing diagnostic, prophylactic and therapeutic measures against the disease. This attempt could be a preliminary step towards the expansion of ML application in future directions.

Bamorovat Mehdi, Sharifi Iraj, Rashedi Esmat, Shafiian Alireza, Sharifi Fatemeh, Khosravi Ahmad, Tahmouresi Amirhossein

2021

Public Health Public Health

Nutritional markers of undiagnosed type 2 diabetes in adults: Findings of a machine learning analysis with external validation and benchmarking.

In PloS one ; h5-index 176.0

OBJECTIVES : Using a nationally-representative, cross-sectional cohort, we examined nutritional markers of undiagnosed type 2 diabetes in adults via machine learning.

METHODS : A total of 16429 men and non-pregnant women ≥ 20 years of age were analysed from five consecutive cycles of the National Health and Nutrition Examination Survey. Cohorts from years 2013-2016 (n = 6673) was used for external validation. Undiagnosed type 2 diabetes was determined by a negative response to the question "Have you ever been told by a doctor that you have diabetes?" and a positive glycaemic response to one or more of the three diagnostic tests (HbA1c > 6.4% or FPG >125 mg/dl or 2-hr post-OGTT glucose > 200mg/dl). Following comprehensive literature search, 114 potential nutritional markers were modelled with 13 behavioural and 12 socio-economic variables. We tested three machine learning algorithms on original and resampled training datasets built using three resampling methods. From this, the derived 12 predictive models were validated on internal- and external validation cohorts. Magnitudes of associations were gauged through odds ratios in logistic models and variable importance in others. Models were benchmarked against the ADA diabetes risk test.

RESULTS : The prevalence of undiagnosed type 2 diabetes was 5.26%. Four best-performing models (AUROC range: 74.9%-75.7%) classified 39 markers of undiagnosed type 2 diabetes; 28 via one or more of the three best-performing non-linear/ensemble models and 11 uniquely by the logistic model. They comprised 14 nutrient-based, 12 anthropometry-based, 9 socio-behavioural, and 4 diet-associated markers. AUROC of all models were on a par with ADA diabetes risk test on both internal and external validation cohorts (p>0.05).

CONCLUSIONS : Models performed comparably to the chosen benchmark. Novel behavioural markers such as the number of meals not prepared from home were revealed. This approach may be useful in nutritional epidemiology to unravel new associations with type 2 diabetes.

De Silva Kushan, Lim Siew, Mousa Aya, Teede Helena, Forbes Andrew, Demmer Ryan T, Jönsson Daniel, Enticott Joanne

2021

General General

Triple sulfur-oxygen-strontium isotopes probabilistic geographic assignment of archaeological remains using a novel sulfur isoscape of western Europe.

In PloS one ; h5-index 176.0

Sulfur isotope composition of organic tissues is a commonly used tool for gathering information about provenance and diet in archaeology and paleoecology. However, the lack of maps predicting sulfur isotope variations on the landscape limits the possibility to use this isotopic system in quantitative geographic assignments. We compiled a database of 2,680 sulfur isotope analyses in the collagen of archaeological human and animal teeth from 221 individual locations across Western Europe. We used this isotopic compilation and remote sensing data to apply a multivariate machine-learning regression, and to predict sulfur isotope variations across Western Europe. The resulting model shows that sulfur isotope patterns are highly predictable, with 65% of sulfur isotope variations explained using only 4 variables representing marine sulfate deposition and local geological conditions. We used this novel sulfur isoscape and existing strontium and oxygen isoscapes of Western Europe to apply triple isotopes continuous-surface probabilistic geographic assignments to assess the origin of a series of teeth from local animals and humans from Brittany. We accurately and precisely constrained the origin of these individuals to limited regions of Brittany. This approach is broadly transferable to studies in archaeology and paleoecology as illustrated in a companion paper (Colleter et al. 2021).

Bataille Clément P, Jaouen Klervia, Milano Stefania, Trost Manuel, Steinbrenner Sven, Crubézy Éric, Colleter Rozenn

2021

General General

The research for the function evaluation of facial nerve and the mechanisms of rehabilitation training.

In Medicine

BACKGROUND : Peripheral facial paralysis (PFP) is a common peripheral neural disease. Acupuncture treatment combined with PFP rehabilitation exercises is a routine method of PFP treatment. This article is to provide a new visual and objective evaluation method for exploring the mechanism and efficacy of acupuncture treatment on PFP, and develop an interactive augmented facial nerve function rehabilitation training system with multiple training models.

METHODS : This prospective and observational trial will recruit 200 eligible participants for the following study. In the trial, the laser speckle contrast analysis (LASCA) technology will be applied to monitor the microcirculation of facial blood flow during acupuncture, and real-time monitoring algorithms, data sampling, and digital imaging methods will be conducted by machine learning and image segmentation. Then, a database of patient facial expressions will be built, the correlation between surface blood flow perfusion volume and facial structure symmetry will be analyzed, combined with scale assessment and electrophysiological detection. In addition, we will also explore the objectivity and effectiveness of LASCA in the evaluation of facial paralysis (FP), and the changes in blood flow microcirculation before and after acupuncture treatment will be analyzed.

RESULTS : The standard image of the facial target area with facial nerve injury will be manually segmented by the convolutional neural network method. The blood flow images of the eyelid, cheek, and mandible of the patients' affected and healthy side will be compared and evaluated. Laser speckle blood flow symmetry Pr and its changes in FP condition evolution and prognosis outcome will be measured, and relevant characteristic signals values will be extracted. Finally, COX regression analysis method is conducted to establish a higher accuracy prediction model of FP with cross-validation based on laser speckle blood flow imaging technology.

CONCLUSIONS : We use modern interdisciplinary high-tech technologies to explore the mechanism of acupuncture rehabilitation training in PFP. And we will provide evidence for the feasibility of using the LASCA technique as a typing diagnosis of FP in the acupuncture rehabilitation treatment of PFP.

REGISTRATION NUMBER : ChiCTR1800019463.

Si-Yi Han, Ling Wang, Hai-Bo Yu, Yan-Hua Gou, Wei-Zheng Zhong, Xing-Xian Huang, Shao-Yun Zhang, Yong-Feng Liu, Yi-Rong Chen

2021-May-07

General General

Integrating Resonant Recognition Model and Stockwell Transform for Localization of Hotspots in Tubulin.

In IEEE transactions on nanobioscience

Tubulin is a promising target for designing anti-cancer drugs. Identification of hotspots in multifunctional Tubulin protein provides insights for new drug discovery. Although machine learning techniques have shown significant results in prediction, they fail to identify the hotspots corresponding to a particular biological function. This paper presents a signal processing technique combining resonant recognition model (RRM) and Stockwell Transform (ST) for the identification of hotspots corresponding to a particular functionality. The characteristic frequency (CF) representing a specific biological function is determined using the RRM. Then the spectrum of the protein sequence is computed using ST. The CF is filtered from the ST spectrum using a time-frequency mask. The energy peaks in the filtered sequence represent the hotspots. The hotspots predicted by the proposed method are compared with the experimentally detected binding residues of Tubulin stabilizing drug Taxol and destabilizing drug Colchicine present in the Tubulin protein. Out of the 53 experimentally identified hotspots, 60% are predicted by the proposed method whereas around 20% are predicted by existing machine learning based methods. Additionally, the proposed method predicts some new hot spots, which may be investigated.

Mahapatra Satyajit, Sahu Sitanshu Sekhar

2021-May-05

General General

Optimal Donor Selection for Hematopoietic Cell Transplantation Using Bayesian Machine Learning.

In JCO clinical cancer informatics

PURPOSE : Donor selection practices for matched unrelated donor (MUD) hematopoietic cell transplantation (HCT) vary, and the impact of optimizing donor selection in a patient-specific way using modern machine learning (ML) models has not been studied.

METHODS : We trained a Bayesian ML model in 10,318 patients who underwent MUD HCT from 1999 to 2014 to provide patient- and donor-specific predictions of clinically severe (grade 3 or 4) acute graft-versus-host disease or death by day 180. The model was validated in 3,501 patients from 2015 to 2016 with archived records of potential donors at search. Donor selection optimizing predicted outcomes was implemented over either an unlimited donor pool or the donors in the search archives. Posterior mean differences in outcomes from optimal donor selection versus actual practice were summarized per patient and across the population with 95% intervals.

RESULTS : Event rates were 33% (training) and 37% (validation). Among donor features, only age affected outcomes, with the effect consistent regardless of patient features. The median (interquartile range) difference in age between the youngest donor at search and the selected donor was 6 (1-10) years, whereas the number of donors per patient younger than the selected donor was 6 (1-36). Fourteen percent of the validation data set had an approximate 5% absolute reduction in event rates from selecting the youngest donor at search versus the actual donor used, leading to an absolute population reduction of 1% (95% interval, 0 to 3).

CONCLUSION : We confirmed the singular importance of selecting the youngest available MUD, irrespective of patient features, identified potential for improved HCT outcomes by selecting a younger MUD, and demonstrated use of novel ML models transferable to optimize other complex treatment decisions in a patient-specific way.

Logan Brent R, Maiers Martin J, Sparapani Rodney A, Laud Purushottam W, Spellman Stephen R, McCulloch Robert E, Shaw Bronwen E

2021-May

General General

Extending Approximate Bayesian Computation with Supervised Machine Learning to infer demographic history from genetic polymorphisms using DIYABC Random Forest.

In Molecular ecology resources

Simulation-based methods such as Approximate Bayesian Computation (ABC) are well-adapted to the analysis of complex scenarios of populations and species genetic history. In this context, supervised machine learning (SML) methods provide attractive statistical solutions to conduct efficient inferences about scenario choice and parameter estimation. The Random Forest methodology (RF) is a powerful ensemble of SML algorithms used for classification or regression problems. RF allows conducting inferences at a low computational cost, without preliminary selection of the relevant components of the ABC summary statistics, and bypassing the derivation of ABC tolerance levels. We have implemented a set of RF algorithms to process inferences using simulated datasets generated from an extended version of the population genetic simulator implemented in DIYABC v2.1.0. The resulting computer package, named DIYABC Random Forest v1.0, integrates two functionalities into a user-friendly interface: the simulation under custom evolutionary scenarios of different types of molecular data (microsatellites, DNA sequences or SNPs) and RF treatments including statistical tools to evaluate the power and accuracy of inferences. We illustrate the functionalities of DIYABC Random Forest v1.0 for both scenario choice and parameter estimation through the analysis of pseudo-observed and real datasets corresponding to pool-sequencing and individual-sequencing SNP datasets. Because of the properties inherent to the implemented RF methods and the large feature vector (including various summary statistics and their linear combinations) available for SNP data, DIYABC Random Forest v1.0 can efficiently contribute to the analysis of large SNP datasets to make inferences about complex population genetic histories.

Collin François-David, Durif Ghislain, Raynal Louis, Lombaert Eric, Gautier Mathieu, Vitalis Renaud, Marin Jean-Michel, Estoup Arnaud

2021-May-05

Approximate Bayesian Computation, Random Forest, Supervised Machine Learning, model or scenario selection, parameter estimation, population genetics

Radiology Radiology

PA-ResSeg: A Phase Attention Residual Network for Liver Tumor Segmentation from Multi-phase CT Images.

In Medical physics ; h5-index 59.0

PURPOSE : Liver tumor segmentation is a crucial prerequisite for computer aided diagnosis of liver tumors. In the clinical diagnosis of liver tumors, radiologists usually examine multi-phase CT images as these images provide abundant and complementary information of tumors. However, most known automatic segmentation methods extract tumor features from CT images merely of a single phase, in which valuable multi-phase information is ignored. Therefore, it is highly demanded to develop a method effectively incorporating multi-phase information for automatic and accurate liver tumor segmentation.

METHODS : In this paper, we propose a phase attention residual network (PA-ResSeg) to model multi-phase features for accurate liver tumor segmentation. A phase attention (PA) is newly proposed to additionally exploit the images of arterial (ART) phase to facilitate the segmentation of portal venous (PV) phase. The PA block consists of an intra-phase attention (Intra-PA) module and an inter-phase attention (Inter-PA) module to capture channel-wise self-dependencies and cross-phase interdependencies, respectively. Thus it enables the network to learn more representative multi-phase features by refining the PV features according to the channel dependencies and recalibrating the ART features based on the learned interdependencies between phases. We propose a PA-based multi-scale fusion (MSF) architecture to embed the PA blocks in the network at multiple levels along the encoding path to fuse multi-scale features from multi-phase images. Moreover, a 3D boundary-enhanced loss (BE-loss) is proposed for training to make the network more sensitive to boundaries.

RESULTS : To evaluate the performance of our proposed PA-ResSeg, we conducted experiments on a multi-phase CT dataset of focal liver lesions (MPCT-FLLs). Experimental results show the effectiveness of the proposed method by achieving a dice per case (DPC) of 0.7787, a dice global (DG) of 0.8682, a volumetric overlap error (VOE) of 0.3328 and a relative volume difference (RVD) of 0.0443 on the MPCT-FLLs. Furthermore, to validate the effectiveness and robustness of PA-ResSeg, we conducted extra experiments on another multi-phase liver tumor dataset and obtained a DPC of 0.8290, a DG of 0.9132, a VOE of 0.2637 and a RVD of 0.0163. The proposed method shows its robustness and generalization capability in different datasets and different backbones.

CONCLUSIONS : The study demonstrates that our method can effectively model information from multi-phase CT images to segment liver tumors and outperforms other state-of-the-art methods. The PA-based MSF method can learn more representative multi-phase features at multiple scales and thereby improve the segmentation performance. Besides, the proposed 3D BE-loss is conducive to tumor boundary segmentation by enforcing the network focus on boundary regions and marginal slices. Experimental results evaluated by quantitative metrics demonstrate the superiority of our PA-ResSeg over the best-known methods.

Xu Yingying, Cai Ming, Lin Lanfen, Zhang Yue, Hu Hongjie, Peng Zhiyi, Zhang Qiaowei, Chen Qingqing, Mao Xiongwei, Iwamoto Yutaro, Han Xian-Hua, Chen Yen-Wei, Tong Ruofeng

2021-May-05

3D boundary-enhanced loss, liver tumor segmentation, multi-phase CT, multi-scale fusion, phase attention

Radiology Radiology

Applying MRI Intensity Normalization on Non-Bone Tissues to Facilitate Pseudo-CT Synthesis from MRI.

In Diagnostics (Basel, Switzerland)

This study aimed to facilitate pseudo-CT synthesis from MRI by normalizing MRI intensity of the same tissue type to a similar intensity level. MRI intensity normalization was conducted through dividing MRI by a shading map, which is a smoothed ratio image between MRI and a three-intensity mask. Regarding pseudo-CT synthesis from MRI, a conversion model based on a three-layer convolutional neural network was trained and validated. Before MRI intensity normalization, the mean value ± standard deviation of fat tissue in 0.35 T chest MRI was 297 ± 73 (coefficient of variation (CV) = 24.58%), which was 533 ± 91 (CV = 17.07%) in 1.5 T abdominal MRI. The corresponding results were 149 ± 32 (CV = 21.48%) and 148 ± 28 (CV = 18.92%) after intensity normalization. With regards to pseudo-CT synthesis from MRI, the differences in mean values between pseudo-CT and real CT were 3, 15, and 12 HU for soft tissue, fat, and lung/air in 0.35 T chest imaging, respectively, while the corresponding results were 3, 14, and 15 HU in 1.5 T abdominal imaging. Overall, the proposed workflow is reliable in pseudo-CT synthesis from MRI and is more practicable in clinical routine practice compared with deep learning methods, which demand a high level of resources for building a conversion model.

Hou Kuei-Yuan, Lu Hao-Yuan, Yang Ching-Ching

2021-Apr-30

MRI intensity normalization, convolutional neural network, pseudo-CT synthesis

General General

Personalized analytics and wearable biosensor platform for early detection of COVID-19 decompensation (DeCODe: Detection of COVID-19 Decompensation): protocol for development of COVID-19 Decompensation Index.

In JMIR research protocols ; h5-index 26.0

BACKGROUND : During the COVID-19 pandemic, novel digital health technologies have the potential to improve our understanding of SARS Co-V2/ COVID-19, improve care delivery and produce better health outcomes. The National Institutes of Health called on digital health leaders to contribute to a high-quality data repository that will support the work of researchers to make discoveries not possible through small, limited data sets.

OBJECTIVE : To this end, we seek to develop a COVID-19 digital biomarker that could provide early detection of a patient's physiologic worsening or decompensation. We propose developing and validating a COVID-19 Decompensation Index (CDI) in a two-phased project that builds off existing wearable biosensor-derived analytics generated by physIQ's end-to-end cloud platform for continuous monitoring of physiology with wearable biosensors. This effort will achieve two primary objectives: 1) collect adequate data to enable the development of the CDI; and 2) collect rich deidentified clinical data correlative with outcomes and symptomology related to COVID-19 disease progression. Secondary objectives include evaluation of feasibility and usability of pinpointIQ™, the digital platform through which data is gathered, analyzed, and displayed.

METHODS : This study is a prospective, non-randomized, open-label, two-phase design. Phase I will involve data collection for the NIH digital data hub as well as data to support the preliminary development of the CDI. Phase II will involve data collection for the hub and contribute to continued refinement and validation of the CDI. While this study will focus on development of a CDI, the digital platform will also be evaluated for feasibility and usability while clinicians deliver care to continuously monitored patients enrolled in the study.

RESULTS : Our target COVID-19 Decompensation Index (CDI) will be a binary classifier trained to distinguish between participants decompensating and not decompensating. The primary performance metric for CDI will be ROC AUC with a minimum performance criterion of AUC ≥ 0.75 (significance α = 0.05 and power 1 - β = 0.80). Determination of sex/gender, race or ethnic characteristics that impact differences in the CDI performance, as well as lead time-time to predict decompensation and the relationship to ultimate severity of disease based on the World Health Organization COVID-19 Ordinal Scale will be explored.

CONCLUSIONS : Using machine learning techniques on a large data set of COVID-19 positive patients could produce valuable insights into the physiology of COVID-19 as well as a digital biomarker for COVID-19 decompensation. We plan, with this study, to develop a tool that can uniquely reflect the physiologic data of a diverse population and contribute to a trove of high-quality data that will help researchers better understand COVID-19.

CLINICALTRIAL : Trial Registration: ClinicalTrials.gov NCT NCT04575532.

Larimer Karen, Wegerich Stephan, Splan Joel, Chestek David, Prendergast Heather, Vanden Hoek Terry

2021-May-04

General General

Convolutional neural networks for the diagnosis and prognosis of the coronavirus disease pandemic.

In Visual computing for industry, biomedicine, and art

A neural network is one of the current trends in deep learning, which is increasingly gaining attention owing to its contribution in transforming the different facets of human life. It also paves a way to approach the current crisis caused by the coronavirus disease (COVID-19) from all scientific directions. Convolutional neural network (CNN), a type of neural network, is extensively applied in the medical field, and is particularly useful in the current COVID-19 pandemic. In this article, we present the application of CNNs for the diagnosis and prognosis of COVID-19 using X-ray and computed tomography (CT) images of COVID-19 patients. The CNN models discussed in this review were mainly developed for the detection, classification, and segmentation of COVID-19 images. The base models used for detection and classification were AlexNet, Visual Geometry Group Network with 16 layers, residual network, DensNet, GoogLeNet, MobileNet, Inception, and extreme Inception. U-Net and voxel-based broad learning network were used for segmentation. Even with limited datasets, these methods proved to be beneficial for efficiently identifying the occurrence of COVID-19. To further validate these observations, we conducted an experimental study using a simple CNN framework for the binary classification of COVID-19 CT images. We achieved an accuracy of 93% with an F1-score of 0.93. Thus, with the availability of improved medical image datasets, it is evident that CNNs are very useful for the efficient diagnosis and prognosis of COVID-19.

Kugunavar Sneha, Prabhakar C J

2021-May-05

COVID-19, Convolutional neural network, Deep learning, Medical image analysis, Neural network

General General

Development of intra-oral automated landmark recognition (ALR) for dental and occlusal outcome measurements.

In European journal of orthodontics

BACKGROUND : Previous studies embracing digital technology and automated methods of scoring dental arch relationships have shown that such technology is valid and accurate. To date, however there is no published literature on artificial intelligence and machine learning to completely automate the process of dental landmark recognition.

OBJECTIVES : This study aimed to develop and evaluate a fully automated system and software tool for the identification of landmarks on human teeth using geometric computing, image segmenting, and machine learning technology.

METHODS : Two hundred and thirty-nine digital models were used in the automated landmark recognition (ALR) validation phase, 161 of which were digital models from cleft palate subjects aged 5 years. These were manually annotated to facilitate qualitative validation. Additionally, landmarks were placed on 20 adult digital models manually by 3 independent observers. The same models were subjected to scoring using the ALR software and the differences (in mm) were calculated. All the teeth from the 239 models were evaluated for correct recognition by the ALR with a breakdown to find which stages of the process caused the errors.

RESULTS : The results revealed that 1526 out of 1915 teeth (79.7%) were correctly identified, and the accuracy validation gave 95% confidence intervals for the geometric mean error of [0.285, 0.317] for the humans and [0.269, 0.325] for ALR-a negligible difference.

CONCLUSIONS/IMPLICATIONS : It is anticipated that ALR software tool will have applications throughout clinical dentistry and anthropology, and in research will constitute an accurate and objective tool for handling large datasets without the need for time intensive employment of experts to place landmarks manually.

Woodsend Brénainn, Koufoudaki Eirini, Lin Ping, McIntyre Grant, El-Angbawi Ahmed, Aziz Azad, Shaw William, Semb Gunvor, Reesu Gowri Vijay, Mossey Peter A

2021-May-05

Ophthalmology Ophthalmology

Automated Explainable Multidimensional Deep Learning Platform of Retinal Images for Retinopathy of Prematurity Screening.

In JAMA network open

Importance : A retinopathy of prematurity (ROP) diagnosis currently relies on indirect ophthalmoscopy assessed by experienced ophthalmologists. A deep learning algorithm based on retinal images may facilitate early detection and timely treatment of ROP to improve visual outcomes.

Objective : To develop a retinal image-based, multidimensional, automated, deep learning platform for ROP screening and validate its performance accuracy.

Design, Setting, and Participants : A total of 14 108 eyes of 8652 preterm infants who received ROP screening from 4 centers from November 4, 2010, to November 14, 2019, were included, and a total of 52 249 retinal images were randomly split into training, validation, and test sets. Four main dimensional independent classifiers were developed, including image quality, any stage of ROP, intraocular hemorrhage, and preplus/plus disease. Referral-warranted ROP was automatically generated by integrating the results of 4 classifiers at the image, eye, and patient levels. DeepSHAP, a method based on DeepLIFT and Shapley values (solution concepts in cooperative game theory), was adopted as the heat map technology to explain the predictions. The performance of the platform was further validated as compared with that of the experienced ROP experts. Data were analyzed from February 12, 2020, to June 24, 2020.

Exposure : A deep learning algorithm.

Main Outcomes and Measures : The performance of each classifier included true negative, false positive, false negative, true positive, F1 score, sensitivity, specificity, receiver operating characteristic, area under curve (AUC), and Cohen unweighted κ.

Results : A total of 14 108 eyes of 8652 preterm infants (mean [SD] gestational age, 32.9 [3.1] weeks; 4818 boys [60.4%] of 7973 with known sex) received ROP screening. The performance of all classifiers achieved an F1 score of 0.718 to 0.981, a sensitivity of 0.918 to 0.982, a specificity of 0.949 to 0.992, and an AUC of 0.983 to 0.998, whereas that of the referral system achieved an F1 score of 0.898 to 0.956, a sensitivity of 0.981 to 0.986, a specificity of 0.939 to 0.974, and an AUC of 0.9901 to 0.9956. Fine-grained and class-discriminative heat maps were generated by DeepSHAP in real time. The platform achieved a Cohen unweighted κ of 0.86 to 0.98 compared with a Cohen κ of 0.93 to 0.98 by the ROP experts.

Conclusions and Relevance : In this diagnostic study, an automated ROP screening platform was able to identify and classify multidimensional pathologic lesions in the retinal images. This platform may be able to assist routine ROP screening in general and children hospitals.

Wang Ji, Ji Jie, Zhang Mingzhi, Lin Jian-Wei, Zhang Guihua, Gong Weifen, Cen Ling-Ping, Lu Yamei, Huang Xuelin, Huang Dingguo, Li Taiping, Ng Tsz Kin, Pang Chi Pui

2021-May-03

General General

LitSuggest: a web-based system for literature recommendation and curation using machine learning.

In Nucleic acids research ; h5-index 217.0

Searching and reading relevant literature is a routine practice in biomedical research. However, it is challenging for a user to design optimal search queries using all the keywords related to a given topic. As such, existing search systems such as PubMed often return suboptimal results. Several computational methods have been proposed as an effective alternative to keyword-based query methods for literature recommendation. However, those methods require specialized knowledge in machine learning and natural language processing, which can make them difficult for biologists to utilize. In this paper, we propose LitSuggest, a web server that provides an all-in-one literature recommendation and curation service to help biomedical researchers stay up to date with scientific literature. LitSuggest combines advanced machine learning techniques for suggesting relevant PubMed articles with high accuracy. In addition to innovative text-processing methods, LitSuggest offers multiple advantages over existing tools. First, LitSuggest allows users to curate, organize, and download classification results in a single interface. Second, users can easily fine-tune LitSuggest results by updating the training corpus. Third, results can be readily shared, enabling collaborative analysis and curation of scientific literature. Finally, LitSuggest provides an automated personalized weekly digest of newly published articles for each user's project. LitSuggest is publicly available at https://www.ncbi.nlm.nih.gov/research/litsuggest.

Allot Alexis, Lee Kyubum, Chen Qingyu, Luo Ling, Lu Zhiyong

2021-May-05

General General

A self-attention model for inferring cooperativity between regulatory features.

In Nucleic acids research ; h5-index 217.0

Deep learning has demonstrated its predictive power in modeling complex biological phenomena such as gene expression. The value of these models hinges not only on their accuracy, but also on the ability to extract biologically relevant information from the trained models. While there has been much recent work on developing feature attribution methods that discover the most important features for a given sequence, inferring cooperativity between regulatory elements, which is the hallmark of phenomena such as gene expression, remains an open problem. We present SATORI, a Self-ATtentiOn based model to detect Regulatory element Interactions. Our approach combines convolutional layers with a self-attention mechanism that helps us capture a global view of the landscape of interactions between regulatory elements in a sequence. A comprehensive evaluation demonstrates the ability of SATORI to identify numerous statistically significant TF-TF interactions, many of which have been previously reported. Our method is able to detect higher numbers of experimentally verified TF-TF interactions than existing methods, and has the advantage of not requiring a computationally expensive post-processing step. Finally, SATORI can be used for detection of any type of feature interaction in models that use a similar attention mechanism, and is not limited to the detection of TF-TF interactions.

Ullah Fahad, Ben-Hur Asa

2021-May-05

General General

Estimation of trapezoidal-shaped overlapping nuclear pulse parameters based on a deep learning CNN-LSTM model.

In Journal of synchrotron radiation

The Long Short-Term Memory neural network (LSTM) has excellent learning ability for the time series of the nuclear pulse signal. It can accurately estimate the parameters (such as amplitude, time constant, etc.) of the digitally shaped nuclear pulse signal (especially the overlapping pulse signal). However, due to the large number of pulse sequences, the direct use of these sequences as samples to train the LSTM increases the complexity of the network, resulting in a lower training efficiency of the model. The convolution neural network (CNN) can effectively extract the sequence samples by using its unique convolution kernel structure, thus greatly reducing the number of sequence samples. Therefore, the CNN-LSTM deep neural network is used to estimate the parameters of overlapping pulse signals after digital trapezoidal shaping of exponential signals. Firstly, the estimation of the trapezoidal overlapping nuclear pulse is considered to be obtained after the superposition of multiple exponential nuclear pulses followed by trapezoidal shaping. Then, a data set containing multiple samples is set up; each sample is composed of the sequence of sampling values of the trapezoidal overlapping nuclear pulse and the set of shaping parameters of the exponential pulse before digital shaping. Secondly, the CNN is used to extract the abstract features of the training set in these samples, and then these abstract features are applied to the training of the LSTM model. In the training process, the pulse parameter set estimated by the present neural network is calculated by forward propagation. Thirdly, the loss function is used to calculate the loss value between the estimated pulse parameter set and the actual pulse parameter set. Finally, a gradient-based optimization algorithm is applied to update the weight by getting back the loss value together with the gradient of the loss function to the network, so as to realize the purpose of training the network. After model training was completed, the sampled values of the trapezoidal overlapping nuclear pulse were used as input to the CNN-LSTM model to obtain the required parameter set from the output of the CNN-LSTM model. The experimental results show that this method can effectively overcome the shortcomings of local convergence of traditional methods and greatly save the time of model training. At the same time, it can accurately estimate multiple trapezoidal overlapping pulses due to the wide width of the flat top, thus realizing the optimal estimation of nuclear pulse parameters in a global sense, which is a good pulse parameter estimation method.

Ma Xing Ke, Huang Hong Quan, Ji Xiao, Dai He Ye, Wu Jun Hong, Zhao Jing, Yang Fei, Tang Lin, Jiang Kai Ming, Ding Wei Cheng, Zhou Wei

2021-May-01

CNN-LSTM, deep learning, nuclear pulse, trapezoidal shaping

General General

Personalized analytics and wearable biosensor platform for early detection of COVID-19 decompensation (DeCODe: Detection of COVID-19 Decompensation): protocol for development of COVID-19 Decompensation Index.

In JMIR research protocols ; h5-index 26.0

BACKGROUND : During the COVID-19 pandemic, novel digital health technologies have the potential to improve our understanding of SARS Co-V2/ COVID-19, improve care delivery and produce better health outcomes. The National Institutes of Health called on digital health leaders to contribute to a high-quality data repository that will support the work of researchers to make discoveries not possible through small, limited data sets.

OBJECTIVE : To this end, we seek to develop a COVID-19 digital biomarker that could provide early detection of a patient's physiologic worsening or decompensation. We propose developing and validating a COVID-19 Decompensation Index (CDI) in a two-phased project that builds off existing wearable biosensor-derived analytics generated by physIQ's end-to-end cloud platform for continuous monitoring of physiology with wearable biosensors. This effort will achieve two primary objectives: 1) collect adequate data to enable the development of the CDI; and 2) collect rich deidentified clinical data correlative with outcomes and symptomology related to COVID-19 disease progression. Secondary objectives include evaluation of feasibility and usability of pinpointIQ™, the digital platform through which data is gathered, analyzed, and displayed.

METHODS : This study is a prospective, non-randomized, open-label, two-phase design. Phase I will involve data collection for the NIH digital data hub as well as data to support the preliminary development of the CDI. Phase II will involve data collection for the hub and contribute to continued refinement and validation of the CDI. While this study will focus on development of a CDI, the digital platform will also be evaluated for feasibility and usability while clinicians deliver care to continuously monitored patients enrolled in the study.

RESULTS : Our target COVID-19 Decompensation Index (CDI) will be a binary classifier trained to distinguish between participants decompensating and not decompensating. The primary performance metric for CDI will be ROC AUC with a minimum performance criterion of AUC ≥ 0.75 (significance α = 0.05 and power 1 - β = 0.80). Determination of sex/gender, race or ethnic characteristics that impact differences in the CDI performance, as well as lead time-time to predict decompensation and the relationship to ultimate severity of disease based on the World Health Organization COVID-19 Ordinal Scale will be explored.

CONCLUSIONS : Using machine learning techniques on a large data set of COVID-19 positive patients could produce valuable insights into the physiology of COVID-19 as well as a digital biomarker for COVID-19 decompensation. We plan, with this study, to develop a tool that can uniquely reflect the physiologic data of a diverse population and contribute to a trove of high-quality data that will help researchers better understand COVID-19.

CLINICALTRIAL : Trial Registration: ClinicalTrials.gov NCT NCT04575532.

Larimer Karen, Wegerich Stephan, Splan Joel, Chestek David, Prendergast Heather, Vanden Hoek Terry

2021-May-04

General General

Extracting Drug Names and Associated Attributes From Discharge Summaries: Text Mining Study.

In JMIR medical informatics ; h5-index 23.0

BACKGROUND : Drug prescriptions are often recorded in free-text clinical narratives; making this information available in a structured form is important to support many health-related tasks. Although several natural language processing (NLP) methods have been proposed to extract such information, many challenges remain.

OBJECTIVE : This study evaluates the feasibility of using NLP and deep learning approaches for extracting and linking drug names and associated attributes identified in clinical free-text notes and presents an extensive error analysis of different methods. This study initiated with the participation in the 2018 National NLP Clinical Challenges (n2c2) shared task on adverse drug events and medication extraction.

METHODS : The proposed system (DrugEx) consists of a named entity recognizer (NER) to identify drugs and associated attributes and a relation extraction (RE) method to identify the relations between them. For NER, we explored deep learning-based approaches (ie, bidirectional long-short term memory with conditional random fields [BiLSTM-CRFs]) with various embeddings (ie, word embedding, character embedding [CE], and semantic-feature embedding) to investigate how different embeddings influence the performance. A rule-based method was implemented for RE and compared with a context-aware long-short term memory (LSTM) model. The methods were trained and evaluated using the 2018 n2c2 shared task data.

RESULTS : The experiments showed that the best model (BiLSTM-CRFs with pretrained word embeddings [PWE] and CE) achieved lenient micro F-scores of 0.921 for NER, 0.927 for RE, and 0.855 for the end-to-end system. NER, which relies on the pretrained word and semantic embeddings, performed better on most individual entity types, but NER with PWE and CE had the highest classification efficiency among the proposed approaches. Extracting relations using the rule-based method achieved higher accuracy than the context-aware LSTM for most relations. Interestingly, the LSTM model performed notably better in the reason-drug relations, the most challenging relation type.

CONCLUSIONS : The proposed end-to-end system achieved encouraging results and demonstrated the feasibility of using deep learning methods to extract medication information from free-text data.

Alfattni Ghada, Belousov Maksim, Peek Niels, Nenadic Goran

2021-May-05

discharge summaries, electronic health records, information extraction, medication prescriptions, natural language processing

General General

Predicting Intensive Care Unit Length of Stay and Mortality Using Patient Vital Signs: Machine Learning Model Development and Validation.

In JMIR medical informatics ; h5-index 23.0

BACKGROUND : Patient monitoring is vital in all stages of care. In particular, intensive care unit (ICU) patient monitoring has the potential to reduce complications and morbidity, and to increase the quality of care by enabling hospitals to deliver higher-quality, cost-effective patient care, and improve the quality of medical services in the ICU.

OBJECTIVE : We here report the development and validation of ICU length of stay and mortality prediction models. The models will be used in an intelligent ICU patient monitoring module of an Intelligent Remote Patient Monitoring (IRPM) framework that monitors the health status of patients, and generates timely alerts, maneuver guidance, or reports when adverse medical conditions are predicted.

METHODS : We utilized the publicly available Medical Information Mart for Intensive Care (MIMIC) database to extract ICU stay data for adult patients to build two prediction models: one for mortality prediction and another for ICU length of stay. For the mortality model, we applied six commonly used machine learning (ML) binary classification algorithms for predicting the discharge status (survived or not). For the length of stay model, we applied the same six ML algorithms for binary classification using the median patient population ICU stay of 2.64 days. For the regression-based classification, we used two ML algorithms for predicting the number of days. We built two variations of each prediction model: one using 12 baseline demographic and vital sign features, and the other based on our proposed quantiles approach, in which we use 21 extra features engineered from the baseline vital sign features, including their modified means, standard deviations, and quantile percentages.

RESULTS : We could perform predictive modeling with minimal features while maintaining reasonable performance using the quantiles approach. The best accuracy achieved in the mortality model was approximately 89% using the random forest algorithm. The highest accuracy achieved in the length of stay model, based on the population median ICU stay (2.64 days), was approximately 65% using the random forest algorithm.

CONCLUSIONS : The novelty in our approach is that we built models to predict ICU length of stay and mortality with reasonable accuracy based on a combination of ML and the quantiles approach that utilizes only vital signs available from the patient's profile without the need to use any external features. This approach is based on feature engineering of the vital signs by including their modified means, standard deviations, and quantile percentages of the original features, which provided a richer dataset to achieve better predictive power in our models.

Alghatani Khalid, Ammar Nariman, Rezgui Abdelmounaam, Shaban-Nejad Arash

2021-May-05

ICU patient monitoring, clinical intelligence, intensive care unit (ICU), machine learning, predictive model, vital signs measurements

General General

Pain Assessment Tool With Electrodermal Activity for Postoperative Patients: Method Validation Study.

In JMIR mHealth and uHealth

BACKGROUND : Accurate, objective pain assessment is required in the health care domain and clinical settings for appropriate pain management. Automated, objective pain detection from physiological data in patients provides valuable information to hospital staff and caregivers to better manage pain, particularly for patients who are unable to self-report. Galvanic skin response (GSR) is one of the physiologic signals that refers to the changes in sweat gland activity, which can identify features of emotional states and anxiety induced by varying pain levels. This study used different statistical features extracted from GSR data collected from postoperative patients to detect their pain intensity. To the best of our knowledge, this is the first work building pain models using postoperative adult patients instead of healthy subjects.

OBJECTIVE : The goal of this study was to present an automatic pain assessment tool using GSR signals to predict different pain intensities in noncommunicative, postoperative patients.

METHODS : The study was designed to collect biomedical data from postoperative patients reporting moderate to high pain levels. We recruited 25 participants aged 23-89 years. First, a transcutaneous electrical nerve stimulation (TENS) unit was employed to obtain patients' baseline data. In the second part, the Empatica E4 wristband was worn by patients while they were performing low-intensity activities. Patient self-report based on the numeric rating scale (NRS) was used to record pain intensities that were correlated with objectively measured data. The labels were down-sampled from 11 pain levels to 5 different pain intensities, including the baseline. We used 2 different machine learning algorithms to construct the models. The mean decrease impurity method was used to find the top important features for pain prediction and improve the accuracy. We compared our results with a previously published research study to estimate the true performance of our models.

RESULTS : Four different binary classification models were constructed using each machine learning algorithm to classify the baseline and other pain intensities (Baseline [BL] vs Pain Level [PL] 1, BL vs PL2, BL vs PL3, and BL vs PL4). Our models achieved higher accuracy for the first 3 pain models than the BioVid paper approach despite the challenges in analyzing real patient data. For BL vs PL1, BL vs PL2, and BL vs PL4, the highest prediction accuracies were achieved when using a random forest classifier (86.0, 70.0, and 61.5, respectively). For BL vs PL3, we achieved an accuracy of 72.1 using a k-nearest-neighbor classifier.

CONCLUSIONS : We are the first to propose and validate a pain assessment tool to predict different pain levels in real postoperative adult patients using GSR signals. We also exploited feature selection algorithms to find the top important features related to different pain intensities.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID) : RR2-10.2196/17783.

Aqajari Seyed Amir Hossein, Cao Rui, Kasaeyan Naeini Emad, Calderon Michael-David, Zheng Kai, Dutt Nikil, Liljeberg Pasi, Salanterä Sanna, Nelson Ariana M, Rahmani Amir M

2021-May-05

electrodermal activity, health monitoring, machine learning, pain assessment, post-op patients, recognition, wearable electronics

Surgery Surgery

Dexamethasone Induces Changes in Osteogenic Differentiation of Human Mesenchymal Stromal Cells via SOX9 and PPARG, but Not RUNX2.

In International journal of molecular sciences ; h5-index 102.0

Despite the huge body of research on osteogenic differentiation and bone tissue engineering, the translation potential of in vitro results still does not match the effort employed. One reason might be that the protocols used for in vitro research have inherent pitfalls. The synthetic glucocorticoid dexamethasone is commonly used in protocols for trilineage differentiation of human bone marrow mesenchymal stromal cells (hBMSCs). However, in the case of osteogenic commitment, dexamethasone has the main pitfall of inhibiting terminal osteoblast differentiation, and its pro-adipogenic effect is well known. In this work, we aimed to clarify the role of dexamethasone in the osteogenesis of hBMSCs, with a particular focus on off-target differentiation. The results showed that dexamethasone does induce osteogenic differentiation by inhibiting SOX9 expression, but not directly through RUNX2 upregulation as it is commonly thought. Rather, PPARG is concomitantly and strongly upregulated, leading to the formation of adipocyte-like cells within osteogenic cultures. Limiting the exposure to dexamethasone to the first week of differentiation did not affect the mineralization potential. Gene expression levels of RUNX2, SOX9, and PPARG were simulated using approximate Bayesian computation based on a simplified theoretical model, which was able to reproduce the observed experimental trends but with a different range of responses, indicating that other factors should be integrated to fully understand how dexamethasone influences cell fate. In summary, this work provides evidence that current in vitro differentiation protocols based on dexamethasone do not represent a good model, and further research is warranted in this field.

Della Bella Elena, Buetti-Dinh Antoine, Licandro Ginevra, Ahmad Paras, Basoli Valentina, Alini Mauro, Stoddart Martin J

2021-Apr-30

MSC, Osteogenesis, approximate Bayesian computation (ABC), gene expression, glucocorticoids, transcription factors

Surgery Surgery

"P3": an adaptive modeling tool for post-COVID-19 restart of surgical services.

In JAMIA open

Objective : To develop a predictive analytics tool that would help evaluate different scenarios and multiple variables for clearance of surgical patient backlog during the COVID-19 pandemic.

Materials and Methods : Using data from 27 866 cases (May 1 2018-May 1 2020) stored in the Johns Hopkins All Children's data warehouse and inputs from 30 operations-based variables, we built mathematical models for (1) time to clear the case backlog (2), utilization of personal protective equipment (PPE), and (3) assessment of overtime needs.

Results : The tool enabled us to predict desired variables, including number of days to clear the patient backlog, PPE needed, staff/overtime needed, and cost for different backlog reduction scenarios.

Conclusions : Predictive analytics, machine learning, and multiple variable inputs coupled with nimble scenario-creation and a user-friendly visualization helped us to determine the most effective deployment of operating room personnel. Operating rooms worldwide can use this tool to overcome patient backlog safely.

Joshi Divya, Jalali Ali, Whipple Todd, Rehman Mohamed, Ahumada Luis M

2021-Apr

COVID-19, decision support, optimization, predictive analytics, surgical backlog

General General

Deep-learning models for lipid nanoparticle-based drug delivery.

In Nanomedicine (London, England)

Background: Early prediction of time-lapse microscopy experiments enables intelligent data management and decision-making. Aim: Using time-lapse data of HepG2 cells exposed to lipid nanoparticles loaded with mRNA for expression of GFP, the authors hypothesized that it is possible to predict in advance whether a cell will express GFP. Methods: The first modeling approach used a convolutional neural network extracting per-cell features at early time points. These features were then combined and explored using either a long short-term memory network (approach 2) or time series feature extraction and gradient boosting machines (approach 3). Results: Accounting for the temporal dynamics significantly improved performance. Conclusion: The results highlight the benefit of accounting for temporal dynamics when studying drug delivery using high-content imaging.

Harrison Philip J, Wieslander Håkan, Sabirsh Alan, Karlsson Johan, Malmsjö Victor, Hellander Andreas, Wählby Carolina, Spjuth Ola

2021-May-05

artificial neural networks, high-content imaging, machine learning, predictive modeling, time-lapse microscopy

General General

Bipolar disorder: an association of body mass index and cingulate gyrus fractional anisotropy not mediated by systemic inflammation.

In Trends in psychiatry and psychotherapy

OBJECTIVE : To investigate body mass index (BMI) associations with white matter fractional anisotropy (FA) and C-reactive protein (CRP) in individuals with bipolar disorder (BD) during euthymia in comparison with a control group of healthy subjects (CTR).

METHODS : The sample consisted of 101 individuals (BD n=35 and CTR n=66). Regions of interest (ROI) were defined through machine learning approach. For each ROI, a regression model tested the association of FA and BMI controlling for covariates. Peripheral levels of CRP were dosed, correlated with BMI and included in a mediational analysis.

RESULTS : BMI predicted FA of the right cingulate gyrus in BD (AdjR²=.312 F(3)=5.537 p=.004; β=-.340 p=.034), while in CTR, there was no association. There was an interaction effect of BMI and BD diagnosis (F(5)=3.5857 p=.012; Fchange=.227 AdjR²=.093; β=-1.093, p=.048). Furthermore, there was a positive correlation between BMI and CRP in both groups (AdjR²=.170 F(3)=7.337 p<.001; β=.364 p=.001), but it did not act as a mediator of the effect on the FA.

CONCLUSION : Higher BMI is associated with right cingulate microstructure in BD, but not in CTR, and this effect could not be explained by an inflammatory mediation only.

Reckziegel Ramiro, Rabelo-da-Ponte Francisco Diego, Feiten Jacson Gabriel, Remus Isadora Bosini, Goi Pedro Domingues, Vianna-Sulzbach Miréia Fortes, Massuda Raffael, Macedo Danielle, de Lucena David, Czepielewski Letícia Sanguinetti, Gama Clarissa Severino

2021-May-04

bipolar disorder, diffusion tensor imaging, obesity, white matter

Radiology Radiology

Improving phase-based conductivity reconstruction by means of deep learning-based denoising of B 1 + phase data for 3T MRI.

In Magnetic resonance in medicine ; h5-index 66.0

PURPOSE : To denoise B 1 + phase using a deep learning method for phase-based in vivo electrical conductivity reconstruction in a 3T MR system.

METHODS : For B 1 + phase deep-learning denoising, a convolutional neural network (U-net) was chosen. Training was performed on data sets from 10 healthy volunteers. Input data were the real and imaginary components of single averaged spin-echo data (SNR = 45), which was used to approximate the B 1 + phase. For label data, multiple signal-averaged spin-echo data (SNR = 128) were used. Testing was performed on in silico and in vivo data. Reconstructed conductivity maps were derived using phase-based conductivity reconstructions. Additionally, we investigated the usability of the network to various SNR levels, imaging contrasts, and anatomical sites (ie, T1 , T2 , and proton density-weighted brain images and proton density-weighted breast images. In addition, conductivity reconstructions from deep learning-based denoised data were compared with conventional image filters, which were used for data denoising in electrical properties tomography (ie, the Gaussian filtering and the Savitzky-Golay filtering).

RESULTS : The proposed deep learning-based denoising approach showed improvement for B 1 + phase for both in silico and in vivo experiments with reduced quantitative error measures compared with other methods. Subsequently, this resulted in an improvement of reconstructed conductivity maps from the denoised B 1 + phase with deep learning.

CONCLUSION : The results suggest that the proposed approach can be used as an alternative preprocessing method to denoise B 1 + maps for phase-based conductivity reconstruction without relying on image filters or signal averaging.

Jung Kyu-Jin, Mandija Stefano, Kim Jun-Hyeong, Ryu Kanghyun, Jung Soozy, Cui Chuanjiang, Kim Soo-Yeon, Park Mina, van den Berg Cornelis A T, Kim Dong-Hyun

2021-May-05

\n \n B\n 1\n +\n \n \n phase, deep learning, denoising, electrical properties tomography, phase-based conductivity reconstruction

General General

Topological data analysis of collective and individual epithelial cells using persistent homology of loops.

In Soft matter

Interacting, self-propelled particles such as epithelial cells can dynamically self-organize into complex multicellular patterns, which are challenging to classify without a priori information. Classically, different phases and phase transitions have been described based on local ordering, which may not capture structural features at larger length scales. Instead, topological data analysis (TDA) determines the stability of spatial connectivity at varying length scales (i.e. persistent homology), and can compare different particle configurations based on the "cost" of reorganizing one configuration into another. Here, we demonstrate a topology-based machine learning approach for unsupervised profiling of individual and collective phases based on large-scale loops. We show that these topological loops (i.e. dimension 1 homology) are robust to variations in particle number and density, particularly in comparison to connected components (i.e. dimension 0 homology). We use TDA to map out phase diagrams for simulated particles with varying adhesion and propulsion, at constant population size as well as when proliferation is permitted. Next, we use this approach to profile our recent experiments on the clustering of epithelial cells in varying growth factor conditions, which are compared to our simulations. Finally, we characterize the robustness of this approach at varying length scales, with sparse sampling, and over time. Overall, we envision TDA will be broadly applicable as a model-agnostic approach to analyze active systems with varying population size, from cytoskeletal motors to motile cells to flocking or swarming animals.

Bhaskar Dhananjay, Zhang William Y, Wong Ian Y

2021-May-05

General General

Deriving accurate molecular indicators of protein synthesis through Raman-based sparse classification.

In The Analyst

Raman spectroscopy has the ability to retrieve molecular information from live biological samples non-invasively through optical means. Coupled with machine learning, it is possible to use this large amount of information to create models that can predict the state of new samples. We study here linear models, whose separation coefficients can be used to interpret which bands are contributing to the discrimination, and compare the performance of principal component analysis coupled with linear discriminant analysis (PCA/LDA), with regularized logistic regression (Lasso). By applying these methods to single-cell measurements for the detection of macrophage activation, we found that PCA/LDA yields poorer performance in classification compared to Lasso, and underestimates the required sample size to reach stable models. Direct use of Lasso (without PCA) also yields more stable models, and provides sparse separation vectors that directly contain the Raman bands most relevant to classification. To further evaluate these sparse vectors, we apply Lasso to a well-defined case where protein synthesis is inhibited, and show that the separating features are consistent with RNA accumulation and protein levels depletion. Surprisingly, when features are selected purely in terms of their classification power (Lasso), they consist mostly of side bands, while typical strong Raman peaks are not present in the discrimination vector. We propose that this occurs because large Raman bands are representative of a wide variety of intracellular molecules and are therefore less suited for accurate classification.

Pavillon Nicolas, Smith Nicholas I

2021-May-04

General General

Deep learning: step forward to high-resolution in vivo shortwave infrared imaging.

In Journal of biophotonics

Shortwave infrared window (SWIR: 1000-1700 nm) represents a major improvement compared to the NIR-I region (700-900 nm) in terms of temporal and spatial resolutions in depths down to 4 mm. SWIR is a fast and cheap alternative to more precise methods such as X-ray and opto-acoustic imaging. Main obstacles in SWIR imaging are the noise and scattering from tissues and skin that reduce the precision of the method.We demonstrate that the combination of SWIR in vivo imaging in the NIRIIb region (1500-1700 nm) with advanced deep learning image analysis allows to overcome these obstacles and making a large step forward to high resolution imaging: it allows to precisely segment vessels from tissues and noise, provides morphological structure of the vessels network, with learned pseudo-3D shape, their relative position, dynamic information of blood vascularization in depth in small animals and distinguish the vessels types: artieries and veins. For demonstration we use neural network IterNet that exploits structural redundancy of the blood vessels, which provides a useful analysis tool for raw SWIR images. This article is protected by copyright. All rights reserved.

Baulin Vladimir A, Usson Yves, Le Guével Xavier

2021-May-04

SWIR, deep learning, fluorescence, microvessels

Dermatology Dermatology

Impact of environmental factors in predicting daily severity scores of atopic dermatitis.

In Clinical and translational allergy

BACKGROUND : Atopic dermatitis (AD) is a chronic inflammatory skin disease that affects 20% of children worldwide. Environmental factors including weather and air pollutants have been shown to be associated with AD symptoms. However, the time-dependent nature of such a relationship has not been adequately investigated. This paper aims to assess whether real-time data on weather and air pollutants can make short-term prediction of AD severity scores.

METHODS : Using longitudinal data from a published panel study of 177 paediatric patients followed up daily for 17 months, we developed a statistical machine learning model to predict daily AD severity scores for individual study participants. Exposures consisted of daily meteorological variables and concentrations of air pollutants, and outcomes were daily recordings of scores for six AD signs. We developed a mixed-effect autoregressive ordinal logistic regression model, validated it in a forward-chaining setting and evaluated the effects of the environmental factors on the predictive performance.

RESULTS : Our model successfully made daily prediction of the AD severity scores, and the predictive performance was not improved by the addition of measured environmental factors. Potential short-term influence of environmental exposures on daily AD severity scores was outweighed by the underlying persistence of preceding scores.

CONCLUSIONS : Our data does not offer enough evidence to support a claim that weather or air pollutants can make short-term prediction of AD signs. Inferences about the magnitude of the effect of environmental factors on AD severity scores require consideration of their time-dependent dynamic nature.

Hurault Guillem, Delorieux Valentin, Kim Young-Min, Ahn Kangmo, Williams Hywel C, Tanaka Reiko J

2021-Apr

atopic dermatitis, environmental factors, longitudinal data, prediction, statistical machine learning

Pathology Pathology

Aiding clinical assessment of neonatal sepsis using hematological analyzer data with machine learning techniques.

In International journal of laboratory hematology ; h5-index 29.0

INTRODUCTION : Early diagnosis and antibiotic administration are essential for reducing sepsis morbidity and mortality; however, diagnosis remains difficult due to complex pathogenesis and presentation. We created a machine learning model for bacterial sepsis identification in the neonatal intensive care unit (NICU) using hematological analyzer data.

METHODS : Hematological analyzer data were gathered from NICU patients up to 48 hours prior to clinical evaluation for bacterial sepsis. Five models, Support Vector Machine, K-nearest-neighbors, Logistic Regression, Random Forest (RF), and Extreme Gradient boosting (XGBoost), were trained on 60 hematological and nine clinical variables for 2357 cases (1692 control, 665 septic). Clinical feature only models (nine variables) were additionally trained and compared with models including hematological variables. Feature importance was used to assess relative contributions of parameters to performance.

RESULTS : The three best performing models were RF, Logistic Regression, and XGBoost. RF achieved an average accuracy of 0.74, AUC-ROC of 0.73, Sensitivity of 0.38, and Specificity of 0.88. Logistic Regression achieved an average accuracy of 0.70, AUC-ROC of 0.74, Sensitivity of 0.62, and Specificity of 0.73. XGBoost achieved an average accuracy of 0.72, AUC-ROC of 0.71, Sensitivity of 0.40, and Specificity of 0.85. All models with hematological variables had significantly stronger performance than models trained on only clinical features. Neutrophil parameters had the highest average feature importance.

CONCLUSIONS : Machine learning models using hematological analyzer data can classify NICU patients as sepsis positive or negative with stronger performance compared to clinical feature only models. Hematological analyzer variables could augment current sepsis classification machine learning algorithms.

Huang Brian, Wang Robin, Masino Aaron J, Obstfeld Amrom E

2021-May-04

cell population data, hematological analyzer, machine learning, neonatal sepsis, neutrophil fluorescence

General General

Autosomal deletion/insertion polymorphisms for global stratification analyses and ancestry origin inferences of different continental populations by machine learning methods.

In Electrophoresis

A lot of population data of 30 deletion/insertion polymorphisms (DIPs) of the Investigator DIPplex kit in different continental populations have been reported. Here, we assessed genetic distributions of these 30 DIPs in different continental populations to pinpoint candidate ancestry informative DIPs. Besides, the effectiveness of machine learning methods for ancestry analysis was explored. Pairwise informativeness (In) values of 30 DIPs revealed that six loci displayed relatively high In values (>0.1) among different continental populations. Besides, more loci showed high population specific divergence (PSD) values in African population. Based on the pairwise In and PSD values of 30 DIPs, 17 DIPs in the Investigator DIPplex kit were selected to ancestry analyses of African, European and East Asian populations. Even though 30 DIPs provided better ancestry resolution of these continental populations based on the results of PCA and population genetic structure, we found that 17 DIPs could also distinguish these continental populations. More importantly, these 17 DIPs possessed more balanced cumulative PSD distributions in these populations. Six machine learning methods were used to perform ancestry analyses of these continental populations based on 17 DIPs. Obtained results revealed that naïve Bayes manifested the greatest performance; whereas, k nearest neighbor showed relatively low performance. To sum up, these machine learning methods, especially for naïve Bayes, could be used as the valuable tool for ancestry analysis. This article is protected by copyright. All rights reserved.

Jin Xiaoye, Liu Yuluo, Zhang Yuanyuan, Li Yongle, Chen Chuanliang, Wang Hongdan

2021-May-05

Ancestry analysis, Continental populations, Deletion/insertion polymorphisms, Machine learning

General General

NeuroCrypt: Machine Learning Over Encrypted Distributed Neuroimaging Data.

In Neuroinformatics

The field of neuroimaging can greatly benefit from building machine learning models to detect and predict diseases, and discover novel biomarkers, but much of the data collected at various organizations and research centers is unable to be shared due to privacy or regulatory concerns (especially for clinical data or rare disorders). In addition, aggregating data across multiple large studies results in a huge amount of duplicated technical debt and the resources required can be challenging or impossible for an individual site to build. Training on the data distributed across organizations can result in models that generalize much better than models trained on data from any of organizations alone. While there are approaches for decentralized sharing, these often do not provide the highest possible guarantees of sample privacy that only cryptography can provide. In addition, such approaches are often focused on probabilistic solutions. In this paper, we propose an approach that leverages the potential of datasets spread among a number of data collecting organizations by performing joint analyses in a secure and deterministic manner when only encrypted data is shared and manipulated. The approach is based on secure multiparty computation which refers to cryptographic protocols that enable distributed computation of a function over distributed inputs without revealing additional information about the inputs. It enables multiple organizations to train machine learning models on their joint data and apply the trained models to encrypted data without revealing their sensitive data to the other parties. In our proposed approach, organizations (or sites) securely collaborate to build a machine learning model as it would have been trained on the aggregated data of all the organizations combined. Importantly, the approach does not require a trusted party (i.e. aggregator), each contributing site plays an equal role in the process, and no site can learn individual data of any other site. We demonstrate effectiveness of the proposed approach, in a range of empirical evaluations using different machine learning algorithms including logistic regression and convolutional neural network models on human structural and functional magnetic resonance imaging datasets.

Senanayake Nipuna, Podschwadt Robert, Takabi Daniel, Calhoun Vince D, Plis Sergey M

2021-May-04

Convolutional neural networks, Logistic regression, Machine learning, Neuroimaging, Privacy, Secure multiparty computation

Radiology Radiology

Radiomics and Artificial Intelligence Analysis with Textural Metrics Extracted by Contrast-Enhanced Mammography in the Breast Lesions Classification.

In Diagnostics (Basel, Switzerland)

The aim of the study was to estimate the diagnostic accuracy of textural features extracted by dual-energy contrast-enhanced mammography (CEM) images, by carrying out univariate and multivariate statistical analyses including artificial intelligence approaches. In total, 80 patients with known breast lesion were enrolled in this prospective study according to regulations issued by the local Institutional Review Board. All patients underwent dual-energy CEM examination in both craniocaudally (CC) and double acquisition of mediolateral oblique (MLO) projections (early and late). The reference standard was pathology from a surgical specimen for malignant lesions and pathology from a surgical specimen or fine needle aspiration cytology, core or Tru-Cut needle biopsy, and vacuum assisted breast biopsy for benign lesions. In total, 104 samples of 80 patients were analyzed. Furthermore, 48 textural parameters were extracted by manually segmenting regions of interest. Univariate and multivariate approaches were performed: non-parametric Wilcoxon-Mann-Whitney test; receiver operating characteristic (ROC), linear classifier (LDA), decision tree (DT), k-nearest neighbors (KNN), artificial neural network (NNET), and support vector machine (SVM) were utilized. A balancing approach and feature selection methods were used. The univariate analysis showed low accuracy and area under the curve (AUC) for all considered features. Instead, in the multivariate textural analysis, the best performance considering the CC view (accuracy (ACC) = 0.75; AUC = 0.82) was reached with a DT trained with leave-one-out cross-variation (LOOCV) and balanced data (with adaptive synthetic (ADASYN) function) and a subset of three robust textural features (MAD, VARIANCE, and LRLGE). The best performance (ACC = 0.77; AUC = 0.83) considering the early-MLO view was reached with a NNET trained with LOOCV and balanced data (with ADASYN function) and a subset of ten robust features (MEAN, MAD, RANGE, IQR, VARIANCE, CORRELATION, RLV, COARSNESS, BUSYNESS, and STRENGTH). The best performance (ACC = 0.73; AUC = 0.82) considering the late-MLO view was reached with a NNET trained with LOOCV and balanced data (with ADASYN function) and a subset of eleven robust features (MODE, MEDIAN, RANGE, RLN, LRLGE, RLV, LZLGE, GLV_GLSZM, ZSV, COARSNESS, and BUSYNESS). Multivariate analyses using pattern recognition approaches, considering 144 textural features extracted from all three mammographic projections (CC, early MLO, and late MLO), optimized by adaptive synthetic sampling and feature selection operations obtained the best results (ACC = 0.87; AUC = 0.90) and showed the best performance in the discrimination of benign and malignant lesions.

Fusco Roberta, Piccirillo Adele, Sansone Mario, Granata Vincenza, Rubulotta Maria Rosaria, Petrosino Teresa, Barretta Maria Luisa, Vallone Paolo, Di Giacomo Raimondo, Esposito Emanuela, Di Bonito Maurizio, Petrillo Antonella

2021-Apr-30

artificial intelligence, breast, contrast-enhanced digital mammography, radiomics

General General

A novel augmented deep transfer learning for classification of COVID-19 and other thoracic diseases from X-rays.

In Neural computing & applications

Deep learning has provided numerous breakthroughs in natural imaging tasks. However, its successful application to medical images is severely handicapped with the limited amount of annotated training data. Transfer learning is commonly adopted for the medical imaging tasks. However, a large covariant shift between the source domain of natural images and target domain of medical images results in poor transfer learning. Moreover, scarcity of annotated data for the medical imaging tasks causes further problems for effective transfer learning. To address these problems, we develop an augmented ensemble transfer learning technique that leads to significant performance gain over the conventional transfer learning. Our technique uses an ensemble of deep learning models, where the architecture of each network is modified with extra layers to account for dimensionality change between the images of source and target data domains. Moreover, the model is hierarchically tuned to the target domain with augmented training data. Along with the network ensemble, we also utilize an ensemble of dictionaries that are based on features extracted from the augmented models. The dictionary ensemble provides an additional performance boost to our method. We first establish the effectiveness of our technique with the challenging ChestXray-14 radiography data set. Our experimental results show more than 50% reduction in the error rate with our method as compared to the baseline transfer learning technique. We then apply our technique to a recent COVID-19 data set for binary and multi-class classification tasks. Our technique achieves 99.49% accuracy for the binary classification, and 99.24% for multi-class classification.

Altaf Fouzia, Islam Syed M S, Janjua Naeem Khalid

2021-Apr-29

COVID-19, Chest radiography, Computer-aided diagnosis, Deep learning, Dictionary learning, Thoracic disease classification, Transfer learning

Pathology Pathology

Deep Learning based Multi-modal Computing with Feature Disentanglement for MRI Image Synthesis

ArXiv Preprint

Purpose: Different Magnetic resonance imaging (MRI) modalities of the same anatomical structure are required to present different pathological information from the physical level for diagnostic needs. However, it is often difficult to obtain full-sequence MRI images of patients owing to limitations such as time consumption and high cost. The purpose of this work is to develop an algorithm for target MRI sequences prediction with high accuracy, and provide more information for clinical diagnosis. Methods: We propose a deep learning based multi-modal computing model for MRI synthesis with feature disentanglement strategy. To take full advantage of the complementary information provided by different modalities, multi-modal MRI sequences are utilized as input. Notably, the proposed approach decomposes each input modality into modality-invariant space with shared information and modality-specific space with specific information, so that features are extracted separately to effectively process the input data. Subsequently, both of them are fused through the adaptive instance normalization (AdaIN) layer in the decoder. In addition, to address the lack of specific information of the target modality in the test phase, a local adaptive fusion (LAF) module is adopted to generate a modality-like pseudo-target with specific information similar to the ground truth. Results: To evaluate the synthesis performance, we verify our method on the BRATS2015 dataset of 164 subjects. The experimental results demonstrate our approach significantly outperforms the benchmark method and other state-of-the-art medical image synthesis methods in both quantitative and qualitative measures. Compared with the pix2pixGANs method, the PSNR improves from 23.68 to 24.8. Conclusion: The proposed method could be effective in prediction of target MRI sequences, and useful for clinical diagnosis and treatment.

Yuchen Fei, Bo Zhan, Mei Hong, Xi Wu, Jiliu Zhou, Yan Wang

2021-05-06

Radiology Radiology

Effect of AI Explanations on Human Perceptions of Patient-Facing AI-Powered Healthcare Systems.

In Journal of medical systems ; h5-index 48.0

Ongoing research efforts have been examining how to utilize artificial intelligence technology to help healthcare consumers make sense of their clinical data, such as diagnostic radiology reports. How to promote the acceptance of such novel technology is a heated research topic. Recent studies highlight the importance of providing local explanations about AI prediction and model performance to help users determine whether to trust AI's predictions. Despite some efforts, limited empirical research has been conducted to quantitatively measure how AI explanations impact healthcare consumers' perceptions of using patient-facing, AI-powered healthcare systems. The aim of this study is to evaluate the effects of different AI explanations on people's perceptions of AI-powered healthcare system. In this work, we designed and deployed a large-scale experiment (N = 3,423) on Amazon Mechanical Turk (MTurk) to evaluate the effects of AI explanations on people's perceptions in the context of comprehending radiology reports. We created four groups based on two factors-the extent of explanations for the prediction (High vs. Low Transparency) and the model performance (Good vs. Weak AI Model)-and randomly assigned participants to one of the four conditions. Participants were instructed to classify a radiology report as describing a normal or abnormal finding, followed by completing a post-study survey to indicate their perceptions of the AI tool. We found that revealing model performance information can promote people's trust and perceived usefulness of system outputs, while providing local explanations for the rationale of a prediction can promote understandability but not necessarily trust. We also found that when model performance is low, the more information the AI system discloses, the less people would trust the system. Lastly, whether human agrees with AI predictions or not and whether the AI prediction is correct or not could also influence the effect of AI explanations. We conclude this paper by discussing implications for designing AI systems for healthcare consumers to interpret diagnostic report.

Zhang Zhan, Genc Yegin, Wang Dakuo, Ahsen Mehmet Eren, Fan Xiangmin

2021-May-04

Artificial intelligence, Decision making, Diagnostic results, Healthcare, Radiology report, Trust

Radiology Radiology

Impact of artificial intelligence support on accuracy and reading time in breast tomosynthesis image interpretation: a multi-reader multi-case study.

In European radiology ; h5-index 62.0

OBJECTIVES : Digital breast tomosynthesis (DBT) increases sensitivity of mammography and is increasingly implemented in breast cancer screening. However, the large volume of images increases the risk of reading errors and reading time. This study aims to investigate whether the accuracy of breast radiologists reading wide-angle DBT increases with the aid of an artificial intelligence (AI) support system. Also, the impact on reading time was assessed and the stand-alone performance of the AI system in the detection of malignancies was compared to the average radiologist.

METHODS : A multi-reader multi-case study was performed with 240 bilateral DBT exams (71 breasts with cancer lesions, 70 breasts with benign findings, 339 normal breasts). Exams were interpreted by 18 radiologists, with and without AI support, providing cancer suspicion scores per breast. Using AI support, radiologists were shown examination-based and region-based cancer likelihood scores. Area under the receiver operating characteristic curve (AUC) and reading time per exam were compared between reading conditions using mixed-models analysis of variance.

RESULTS : On average, the AUC was higher using AI support (0.863 vs 0.833; p = 0.0025). Using AI support, reading time per DBT exam was reduced (p < 0.001) from 41 (95% CI = 39-42 s) to 36 s (95% CI = 35- 37 s). The AUC of the stand-alone AI system was non-inferior to the AUC of the average radiologist (+0.007, p = 0.8115).

CONCLUSIONS : Radiologists improved their cancer detection and reduced reading time when evaluating DBT examinations using an AI reading support system.

KEY POINTS : • Radiologists improved their cancer detection accuracy in digital breast tomosynthesis (DBT) when using an AI system for support, while simultaneously reducing reading time. • The stand-alone breast cancer detection performance of an AI system is non-inferior to the average performance of radiologists for reading digital breast tomosynthesis exams. • The use of an AI support system could make advanced and more reliable imaging techniques more accessible and could allow for more cost-effective breast screening programs with DBT.

van Winkel Suzanne L, Rodríguez-Ruiz Alejandro, Appelman Linda, Gubern-Mérida Albert, Karssemeijer Nico, Teuwen Jonas, Wanders Alexander J T, Sechopoulos Ioannis, Mann Ritse M

2021-May-04

Artificial intelligence (AI), Breast cancer, Digital breast tomosynthesis (DBT), Mammography, Mass screening

Cardiology Cardiology

Leveraging clinical epigenetics in heart failure with preserved ejection fraction: a call for individualized therapies.

In European heart journal ; h5-index 154.0

Described as the 'single largest unmet need in cardiovascular medicine', heart failure with preserved ejection fraction (HFpEF) remains an untreatable disease currently representing 65% of new heart failure diagnoses. HFpEF is more frequent among women and associates with a poor prognosis and unsustainable healthcare costs. Moreover, the variability in HFpEF phenotypes amplifies complexity and difficulties in the approach. In this perspective, unveiling novel molecular targets is imperative. Epigenetic modifications-defined as changes of DNA, histones, and non-coding RNAs (ncRNAs)-represent a molecular framework through which the environment modulates gene expression. Epigenetic signals acquired over the lifetime lead to chromatin remodelling and affect transcriptional programmes underlying oxidative stress, inflammation, dysmetabolism, and maladaptive left ventricular remodelling, all conditions predisposing to HFpEF. The strong involvement of epigenetic signalling in this setting makes the epigenetic information relevant for diagnostic and therapeutic purposes in patients with HFpEF. The recent advances in high-throughput sequencing, computational epigenetics, and machine learning have enabled the identification of reliable epigenetic biomarkers in cardiovascular patients. Contrary to genetic tools, epigenetic biomarkers mirror the contribution of environmental cues and lifestyle changes and their reversible nature offers a promising opportunity to monitor disease states. The growing understanding of chromatin and ncRNAs biology has led to the development of several Food and Drug Administration approved 'epidrugs' (chromatin modifiers, mimics, anti-miRs) able to prevent transcriptional alterations underpinning left ventricular remodelling and HFpEF. In the present review, we discuss the importance of clinical epigenetics as a new tool to be employed for a personalized management of HFpEF.

Hamdani Nazha, Costantino Sarah, Mügge Andreas, Lebeche Djamel, Tschöpe Carsten, Thum Thomas, Paneni Francesco

2021-May-05

Epigenetics • Heart failure • Chromatin changes • Non-coding RNAs • Precision medicine

Surgery Surgery

"P3": an adaptive modeling tool for post-COVID-19 restart of surgical services.

In JAMIA open

Objective : To develop a predictive analytics tool that would help evaluate different scenarios and multiple variables for clearance of surgical patient backlog during the COVID-19 pandemic.

Materials and Methods : Using data from 27 866 cases (May 1 2018-May 1 2020) stored in the Johns Hopkins All Children's data warehouse and inputs from 30 operations-based variables, we built mathematical models for (1) time to clear the case backlog (2), utilization of personal protective equipment (PPE), and (3) assessment of overtime needs.

Results : The tool enabled us to predict desired variables, including number of days to clear the patient backlog, PPE needed, staff/overtime needed, and cost for different backlog reduction scenarios.

Conclusions : Predictive analytics, machine learning, and multiple variable inputs coupled with nimble scenario-creation and a user-friendly visualization helped us to determine the most effective deployment of operating room personnel. Operating rooms worldwide can use this tool to overcome patient backlog safely.

Joshi Divya, Jalali Ali, Whipple Todd, Rehman Mohamed, Ahumada Luis M

2021-Apr

COVID-19, decision support, optimization, predictive analytics, surgical backlog

General General

A modern deep learning framework in robot vision for automated bean leaves diseases detection.

In International journal of intelligent robotics and applications

The bean leaves can be affected by several diseases, such as angular leaf spots and bean rust, which can cause big damage to bean crops and decrease their productivity. Thus, treating these diseases in their early stages can improve the quality and quantity of the product. Recently, several robotic frameworks based on image processing and artificial intelligence have been used to treat these diseases in an automated way. However, incorrect diagnosis of the infected leaf can lead to the use of chemical treatments for normal leaf thereby the issue will not be solved, and the process may be costly and harmful. To overcome these issues, a modern deep learning framework in robot vision for the early detection of bean leaves diseases is proposed. The proposed framework is composed of two primary stages, which detect the bean leaves in the input images and diagnosing the diseases within the detected leaves. The U-Net architecture based on a pre-trained ResNet34 encoder is employed for detecting the bean leaves in the input images captured in uncontrolled environmental conditions. In the classification stage, the performance of five diverse deep learning models (e.g., Densenet121, ResNet34, ResNet50, VGG-16, and VGG-19) is assessed accurately to identify the healthiness of bean leaves. The performance of the proposed framework is evaluated using a challenging and extensive dataset composed of 1295 images of three different classes (e.g., Healthy, Angular Leaf Spot, and Bean Rust). In the binary classification task, the best performance is achieved using the Densenet121 model with a CAR of 98.31%, Sensitivity of 99.03%, Specificity of 96.82%, Precision of 98.45%, F1-Score of 98.74%, and AUC of 100%. The higher CAR of 91.01% is obtained using the same model in the multi-classification task, with less than 2 s per image to produce the final decision.

Abed Sudad H, Al-Waisy Alaa S, Mohammed Hussam J, Al-Fahdawi Shumoos

2021-Apr-30

Bean leaves diseases, Deep learning, ResNet34 model, Robot vision, Transfer learning, U-Net architecture

Public Health Public Health

Clinically Applicable Machine Learning Approaches to Identify Attributes of Chronic Kidney Disease (CKD) for Use in Low-Cost Diagnostic Screening.

In IEEE journal of translational engineering in health and medicine

OBJECTIVE : Chronic kidney disease (CKD) is a major public health concern worldwide. High costs of late-stage diagnosis and insufficient testing facilities can contribute to high morbidity and mortality rates in CKD patients, particularly in less developed countries. Thus, early diagnosis aided by vital parameter analytics using affordable computer-aided diagnosis could not only reduce diagnosis costs but improve patient management and outcomes.

METHODS : In this study, we developed machine learning models using selective key pathological categories to identify clinical test attributes that will aid in accurate early diagnosis of CKD. Such an approach will save time and costs for diagnostic screening. We have also evaluated the performance of several classifiers with k-fold cross-validation on optimized datasets derived using these selected clinical test attributes.

RESULTS : Our results suggest that the optimized datasets with important attributes perform well in diagnosis of CKD using our proposed machine learning models. Furthermore, we evaluated clinical test attributes based on urine and blood tests along with clinical parameters that have low costs of acquisition. The predictive models with the optimized and pathologically categorized attributes set yielded high levels of CKD diagnosis accuracy with random forest (RF) classifier being the best performing.

CONCLUSIONS : Our machine learning approach has yielded effective predictive analytics for CKD screening which can be developed as a resource to facilitate improved CKD screening for enhanced and timely treatment plans.

Rashed-Al-Mahfuz Md, Haque Abedul, Azad Akm, Alyami Salem A, Quinn Julian M W, Moni Mohammad Ali

2021

Attribute selection, chronic kidney disease (CKD), computer-aided diagnosis, explainable AI, machine learning (ML)

Radiology Radiology

Pretreatment CT and PET radiomics predicting rectal cancer patients in response to neoadjuvant chemoradiotherapy.

In Reports of practical oncology and radiotherapy : journal of Greatpoland Cancer Center in Poznan and Polish Society of Radiation Oncology

Background : The purpose of this study was to characterize pre-treatment non-contrast computed tomography (CT) and 18F-fluorodeoxyglucose positron emission tomography (PET) based radiomics signatures predictive of pathological response and clinical outcomes in rectal cancer patients treated with neoadjuvant chemoradiotherapy (NACR T).

Materials and methods : An exploratory analysis was performed using pre-treatment non-contrast CT and PET imaging dataset. The association of tumor regression grade (TRG) and neoadjuvant rectal (NAR) score with pre-treatment CT and PET features was assessed using machine learning algorithms. Three separate predictive models were built for composite features from CT + PET.

Results : The patterns of pathological response were TRG 0 (n = 13; 19.7%), 1 (n = 34; 51.5%), 2 (n = 16; 24.2%), and 3 (n = 3; 4.5%). There were 20 (30.3%) patients with low, 22 (33.3%) with intermediate and 24 (36.4%) with high NAR scores. Three separate predictive models were built for composite features from CT + PET and analyzed separately for clinical endpoints. Composite features with α = 0.2 resulted in the best predictive power using logistic regression. For pathological response prediction, the signature resulted in 88.1% accuracy in predicting TRG 0 vs. TRG 1-3; 91% accuracy in predicting TRG 0-1 vs. TRG 2-3. For the surrogate of DFS and OS, it resulted in 67.7% accuracy in predicting low vs. intermediate vs. high NAR scores.

Conclusion : The pre-treatment composite radiomics signatures were highly predictive of pathological response in rectal cancer treated with NACR T. A larger cohort is warranted for further validation.

Yuan Zhigang, Frazer Marissa, Rishi Anupam, Latifi Kujtim, Tomaszewski Michal R, Moros Eduardo G, Feygelman Vladimir, Felder Seth, Sanchez Julian, Dessureault Sophie, Imanirad Iman, Kim Richard D, Harrison Louis B, Hoffe Sarah E, Zhang Geoffrey G, Frakes Jessica M

2021

CT, PET, neoadjuvant chemoradiation therapy, pathologic response, radiomics, rectal cancer

General General

Using supervised machine learning classifiers to estimate likelihood of participating in clinical trials of a de-identified version of ResearchMatch.

In Journal of clinical and translational science

Introduction : Lack of participation in clinical trials (CTs) is a major barrier for the evaluation of new pharmaceuticals and devices. Here we report the results of the analysis of a dataset from ResearchMatch, an online clinical registry, using supervised machine learning approaches and a deep learning approach to discover characteristics of individuals more likely to show an interest in participating in CTs.

Methods : We trained six supervised machine learning classifiers (Logistic Regression (LR), Decision Tree (DT), Gaussian Naïve Bayes (GNB), K-Nearest Neighbor Classifier (KNC), Adaboost Classifier (ABC) and a Random Forest Classifier (RFC)), as well as a deep learning method, Convolutional Neural Network (CNN), using a dataset of 841,377 instances and 20 features, including demographic data, geographic constraints, medical conditions and ResearchMatch visit history. Our outcome variable consisted of responses showing specific participant interest when presented with specific clinical trial opportunity invitations ('yes' or 'no'). Furthermore, we created four subsets from this dataset based on top self-reported medical conditions and gender, which were separately analysed.

Results : The deep learning model outperformed the machine learning classifiers, achieving an area under the curve (AUC) of 0.8105.

Conclusions : The results show sufficient evidence that there are meaningful correlations amongst predictor variables and outcome variable in the datasets analysed using the supervised machine learning classifiers. These approaches show promise in identifying individuals who may be more likely to participate when offered an opportunity for a clinical trial.

Vazquez Janette, Abdelrahman Samir, Byrne Loretta M, Russell Michael, Harris Paul, Facelli Julio C

2020-Sep-04

Supervised machine learning, clinical trial participation, clinical trial recruitment, convolutional neural network, deep learning

Pathology Pathology

Developing image analysis pipelines of whole-slide images: Pre- and post-processing.

In Journal of clinical and translational science

Deep learning has pushed the scope of digital pathology beyond simple digitization and telemedicine. The incorporation of these algorithms in routine workflow is on the horizon and maybe a disruptive technology, reducing processing time, and increasing detection of anomalies. While the newest computational methods enjoy much of the press, incorporating deep learning into standard laboratory workflow requires many more steps than simply training and testing a model. Image analysis using deep learning methods often requires substantial pre- and post-processing order to improve interpretation and prediction. Similar to any data processing pipeline, images must be prepared for modeling and the resultant predictions need further processing for interpretation. Examples include artifact detection, color normalization, image subsampling or tiling, removal of errant predictions, etc. Once processed, predictions are complicated by image file size - typically several gigabytes when unpacked. This forces images to be tiled, meaning that a series of subsamples from the whole-slide image (WSI) are used in modeling. Herein, we review many of these methods as they pertain to the analysis of biopsy slides and discuss the multitude of unique issues that are part of the analysis of very large images.

Smith Byron, Hermsen Meyke, Lesser Elizabeth, Ravichandar Deepak, Kremers Walter

2020-Aug-27

Image analysis, analysis pipeline, computer vision, data science, deep learning, pathology

Public Health Public Health

Lessons and tips for designing a machine learning study using EHR data.

In Journal of clinical and translational science

Machine learning (ML) provides the ability to examine massive datasets and uncover patterns within data without relying on a priori assumptions such as specific variable associations, linearity in relationships, or prespecified statistical interactions. However, the application of ML to healthcare data has been met with mixed results, especially when using administrative datasets such as the electronic health record. The black box nature of many ML algorithms contributes to an erroneous assumption that these algorithms can overcome major data issues inherent in large administrative healthcare data. As with other research endeavors, good data and analytic design is crucial to ML-based studies. In this paper, we will provide an overview of common misconceptions for ML, the corresponding truths, and suggestions for incorporating these methods into healthcare research while maintaining a sound study design.

Arbet Jaron, Brokamp Cole, Meinzen-Derr Jareen, Trinkley Katy E, Spratt Heidi M

2020-Jul-24

Machine learning, electronic health record, healthcare research, research methodology, translational research

Surgery Surgery

Prenatal Exposure to Endocrine-Disrupting Chemicals and Subsequent Brain Structure Changes Revealed by Voxel-Based Morphometry and Generalized Q-Sampling MRI.

In International journal of environmental research and public health ; h5-index 73.0

Previous studies have indicated that prenatal exposure to endocrine-disrupting chemicals (EDCs) can cause adverse neuropsychiatric disorders in children and adolescents. This study aimed to determine the association between the concentrations of prenatal EDCs and brain structure changes in teenagers by using MRI. We recruited 49 mother-child pairs during the third trimester of pregnancy, and collected and examined the concentration of EDCs-including phthalate esters, perfluorochemicals (PFCs), and heavy metals (lead, arsenic, cadmium, and mercury)-in maternal urine and/or serum. MRI voxel-based morphometry (VBM) and generalized q-sampling imaging (GQI) mapping-including generalized fractional anisotropy (GFA), normalized quantitative anisotropy (NQA), and the isotropic value of the orientation distribution function (ISO)-were obtained in teenagers 13-16 years of age in order to find the association between maternal EDC concentrations and possible brain structure alterations in the teenagers' brains. We found that there are several specific vulnerable brain areas/structures associated with prenatal exposure to EDCs, including decreased focal brain volume, primarily in the frontal lobe; high frontoparietal lobe, temporooccipital lobe and cerebellum; and white matter structural alterations, which showed a negative association with GFA/NQA and a positive association with ISO, primarily in the corpus callosum, external and internal capsules, corona radiata, superior fronto-occipital fasciculus, and superior longitudinal fasciculus. Prenatal exposure to EDCs may be associated with specific brain structure alterations in teenagers.

Shen Chao-Yu, Weng Jun-Cheng, Tsai Jeng-Dau, Su Pen-Hua, Chou Ming-Chih, Wang Shu-Li

2021-Apr-30

generalized q-sampling imaging (GQI), heavy metals, perfluorochemicals (PFCs), phthalate esters, voxel-based morphometry (VBM)

Cardiology Cardiology

Artificial Intelligence Applications for COVID-19 in Intensive Care and Emergency Settings: A Systematic Review.

In International journal of environmental research and public health ; h5-index 73.0

Background: Little is known about the role of artificial intelligence (AI) as a decisive technology in the clinical management of COVID-19 patients. We aimed to systematically review and critically appraise the current evidence on AI applications for COVID-19 in intensive care and emergency settings. Methods: We systematically searched PubMed, Embase, Scopus, CINAHL, IEEE Xplore, and ACM Digital Library databases from inception to 1 October 2020, without language restrictions. We included peer-reviewed original studies that applied AI for COVID-19 patients, healthcare workers, or health systems in intensive care, emergency, or prehospital settings. We assessed predictive modelling studies and critically appraised the methodology and key findings of all other studies. Results: Of fourteen eligible studies, eleven developed prognostic or diagnostic AI predictive models, all of which were assessed to be at high risk of bias. Common pitfalls included inadequate sample sizes, poor handling of missing data, failure to account for censored participants, and weak validation of models. Conclusions: Current AI applications for COVID-19 are not ready for deployment in acute care settings, given their limited scope and poor quality. Our findings underscore the need for improvements to facilitate safe and effective clinical adoption of AI applications, for and beyond the COVID-19 pandemic.

Chee Marcel Lucas, Ong Marcus Eng Hock, Siddiqui Fahad Javaid, Zhang Zhongheng, Lim Shir Lynn, Ho Andrew Fu Wah, Liu Nan

2021-Apr-29

COVID-19, artificial intelligence, critical care, emergency department, intensive care, machine learning

General General

Modeling the geospatial evolution of COVID-19 using spatio-temporal convolutional sequence-to-sequence neural networks

ArXiv Preprint

Europe was hit hard by the COVID-19 pandemic and Portugal was one of the most affected countries, having suffered three waves in the first twelve months. Approximately between Jan 19th and Feb 5th 2021 Portugal was the country in the world with the largest incidence rate, with 14-days incidence rates per 100,000 inhabitants in excess of 1000. Despite its importance, accurate prediction of the geospatial evolution of COVID-19 remains a challenge, since existing analytical methods fail to capture the complex dynamics that result from both the contagion within a region and the spreading of the infection from infected neighboring regions. We use a previously developed methodology and official municipality level data from the Portuguese Directorate-General for Health (DGS), relative to the first twelve months of the pandemic, to compute an estimate of the incidence rate in each location of mainland Portugal. The resulting sequence of incidence rate maps was then used as a gold standard to test the effectiveness of different approaches in the prediction of the spatial-temporal evolution of the incidence rate. Four different methods were tested: a simple cell level autoregressive moving average (ARMA) model, a cell level vector autoregressive (VAR) model, a municipality-by-municipality compartmental SIRD model followed by direct block sequential simulation and a convolutional sequence-to-sequence neural network model based on the STConvS2S architecture. We conclude that the convolutional sequence-to-sequence neural network is the best performing method, when predicting the medium-term future incidence rate, using the available information.

Mário Cardoso, André Cavalheiro, Alexandre Borges, Ana F. Duarte, Amílcar Soares, Maria João Pereira, Nuno J. Nunes, Leonardo Azevedo, Arlindo L. Oliveira

2021-05-06

General General

Rediscovering histology: what is new in endoscopy for inflammatory bowel disease?

In Therapeutic advances in gastroenterology

The potential of endoscopic evaluation in the management of inflammatory bowel diseases (IBD) has undoubtedly grown over the last few years. When dealing with IBD patients, histological remission (HR) is now considered a desirable target along with symptomatic and endoscopic remission, due to its association with better long-term outcomes. Consequently, the ability of endoscopic techniques to reflect microscopic findings in vivo without having to collect biopsies has become of upmost importance. In this context, a more accurate evaluation of inflammatory disease activity and the detection of dysplasia represent two mainstay targets for IBD endoscopists. New diagnostic technologies have been developed, such as dye-less chromoendoscopy, endomicroscopy, and molecular imaging, but their real incorporation in daily practice is not yet well defined. Although dye-chromoendoscopy is still recommended as the gold standard approach in dysplasia surveillance, recent research questioned the superiority of this technique over new advanced dye-less modalities [narrow band imaging (NBI), Fuji intelligent color enhancement (FICE), i-scan, blue light imaging (BLI) and linked color imaging (LCI)]. The endoscopic armamentarium might also be enriched by new video capsule endoscopy for monitoring disease activity, and high expectations are placed on the application of artificial intelligence (AI) systems to reduce operator-subjectivity and inter-observer variability. The goal of this review is to provide an updated insight on contemporary knowledge regarding new endoscopic techniques and devices, with special focus on their role in the assessment of disease activity and colorectal cancer surveillance.

Solitano Virginia, D’Amico Ferdinando, Allocca Mariangela, Fiorino Gionata, Zilli Alessandra, Loy Laura, Gilardi Daniela, Radice Simona, Correale Carmen, Danese Silvio, Peyrin-Biroulet Laurent, Furfaro Federica

2021

artificial intelligence, capsule enteroscopy, confocal laser endomicroscopy, dye-chromoendoscopy, endocytoscopy, inflammatory bowel diseases, molecular imaging, virtual chromoendoscopy

General General

Responsible Artificial Intelligence (AI) for Value Formation and Market Performance in Healthcare: the Mediating Role of Patient's Cognitive Engagement.

In Information systems frontiers : a journal of research and innovation

The Healthcare sector has been at the forefront of the adoption of artificial intelligence (AI) technologies. Owing to the nature of the services and the vulnerability of a large section of end-users, the topic of responsible AI has become the subject of widespread study and discussion. We conduct a mixed-method study to identify the constituents of responsible AI in the healthcare sector and investigate its role in value formation and market performance. The study context is India, where AI technologies are in the developing phase. The results from 12 in-depth interviews enrich the more nuanced understanding of how different facets of responsible AI guide healthcare firms in evidence-based medicine and improved patient centered care. PLS-SEM analysis of 290 survey responses validates the theoretical framework and establishes responsible AI as a third-order factor. The 174 dyadic data findings also confirm the mediation mechanism of the patient's cognitive engagement with responsible AI-solutions and perceived value, which leads to market performance.

Kumar Pradeep, Dwivedi Yogesh K, Anand Ambuj

2021-Apr-29

Artificial intelligence, Cognitive engagement, Healthcare, Market performance, Responsible AI, Value formation

General General

Evaluating Methods for Imputing Missing Data from Longitudinal Monitoring of Athlete Workload.

In Journal of sports science & medicine

Missing data can influence calculations of accumulated athlete workload. The objectives were to identify the best single imputation methods and examine workload trends using multiple imputation. External (jumps per hour) and internal (rating of perceived exertion; RPE) workload were recorded for 93 (45 females, 48 males) high school basketball players throughout a season. Recorded data were simulated as missing and imputed using ten imputation methods based on the context of the individual, team and session. Both single imputation and machine learning methods were used to impute the simulated missing data. The difference between the imputed data and the actual workload values was computed as root mean squared error (RMSE). A generalized estimating equation determined the effect of imputation method on RMSE. Multiple imputation of the original dataset, with all known and actual missing workload data, was used to examine trends in longitudinal workload data. Following multiple imputation, a Pearson correlation evaluated the longitudinal association between jump count and sRPE over the season. A single imputation method based on the specific context of the session for which data are missing (team mean) was only outperformed by methods that combine information about the session and the individual (machine learning models). There was a significant and strong association between jump count and sRPE in the original data and imputed datasets using multiple imputation. The amount and nature of the missing data should be considered when choosing a method for single imputation of workload data in youth basketball. Multiple imputation using several predictor variables in a regression model can be used for analyses where workload is accumulated across an entire season.

Benson Lauren C, Stilling Carlyn, Owoeye Oluwatoyosi B A, Emery Carolyn A

2021-Jun

Jump count, basketball, imputation, machine learning, training load

Pathology Pathology

Artificial intelligence and capsule endoscopy: unravelling the future.

In Annals of gastroenterology

The applicability of artificial intelligence (AI) in gastroenterology is a hot topic because of its disruptive nature. Capsule endoscopy plays an important role in several areas of digestive pathology, namely in the investigation of obscure hemorrhagic lesions and the management of inflammatory bowel disease. Therefore, there is growing interest in the use of AI in capsule endoscopy. Several studies have demonstrated the enormous potential of using convolutional neural networks in various areas of capsule endoscopy. The exponential development of the usefulness of AI in capsule endoscopy requires consideration of its medium- and long-term impact on clinical practice. Indeed, the advent of deep learning in the field of capsule endoscopy, with its evolutionary character, could lead to a paradigm shift in clinical activity in this setting. In this review, we aim to illustrate the state of the art of AI in the field of capsule endoscopy.

Mascarenhas Miguel, Afonso João, Andrade Patrícia, Cardoso Hélder, Macedo Guilherme

2021

Capsule endoscopy, artificial intelligence, deep learning, gastroenterology, machine learning

General General

A novel augmented deep transfer learning for classification of COVID-19 and other thoracic diseases from X-rays.

In Neural computing & applications

Deep learning has provided numerous breakthroughs in natural imaging tasks. However, its successful application to medical images is severely handicapped with the limited amount of annotated training data. Transfer learning is commonly adopted for the medical imaging tasks. However, a large covariant shift between the source domain of natural images and target domain of medical images results in poor transfer learning. Moreover, scarcity of annotated data for the medical imaging tasks causes further problems for effective transfer learning. To address these problems, we develop an augmented ensemble transfer learning technique that leads to significant performance gain over the conventional transfer learning. Our technique uses an ensemble of deep learning models, where the architecture of each network is modified with extra layers to account for dimensionality change between the images of source and target data domains. Moreover, the model is hierarchically tuned to the target domain with augmented training data. Along with the network ensemble, we also utilize an ensemble of dictionaries that are based on features extracted from the augmented models. The dictionary ensemble provides an additional performance boost to our method. We first establish the effectiveness of our technique with the challenging ChestXray-14 radiography data set. Our experimental results show more than 50% reduction in the error rate with our method as compared to the baseline transfer learning technique. We then apply our technique to a recent COVID-19 data set for binary and multi-class classification tasks. Our technique achieves 99.49% accuracy for the binary classification, and 99.24% for multi-class classification.

Altaf Fouzia, Islam Syed M S, Janjua Naeem Khalid

2021-Apr-29

COVID-19, Chest radiography, Computer-aided diagnosis, Deep learning, Dictionary learning, Thoracic disease classification, Transfer learning

Ophthalmology Ophthalmology

Deep learning for gradability classification of handheld, non-mydriatic retinal images.

In Scientific reports ; h5-index 158.0

Screening effectively identifies patients at risk of sight-threatening diabetic retinopathy (STDR) when retinal images are captured through dilated pupils. Pharmacological mydriasis is not logistically feasible in non-clinical, community DR screening, where acquiring gradable retinal images using handheld devices exhibits high technical failure rates, reducing STDR detection. Deep learning (DL) based gradability predictions at acquisition could prompt device operators to recapture insufficient quality images, increasing gradable image proportions and consequently STDR detection. Non-mydriatic retinal images were captured as part of SMART India, a cross-sectional, multi-site, community-based, house-to-house DR screening study between August 2018 and December 2019 using the Zeiss Visuscout 100 handheld camera. From 18,277 patient eyes (40,126 images), 16,170 patient eyes (35,319 images) were eligible and 3261 retinal images (1490 patient eyes) were sampled then labelled by two ophthalmologists. Compact DL model area under the receiver operator characteristic curve was 0.93 (0.01) following five-fold cross-validation. Compact DL model agreement (Kappa) were 0.58, 0.69 and 0.69 for high specificity, balanced sensitivity/specificity and high sensitivity operating points compared to an inter-grader agreement of 0.59. Compact DL gradability model performance was favourable compared to ophthalmologists. Compact DL models can effectively classify non-mydriatic, handheld retinal image gradability with potential applications within community-based DR screening.

Nderitu Paul, do Rio Joan M Nunez, Rasheed Rajna, Raman Rajiv, Rajalakshmi Ramachandran, Bergeles Christos, Sivaprasad Sobha

2021-May-04

General General

Weakly supervised temporal model for prediction of breast cancer distant recurrence.

In Scientific reports ; h5-index 158.0

Efficient prediction of cancer recurrence in advance may help to recruit high risk breast cancer patients for clinical trial on-time and can guide a proper treatment plan. Several machine learning approaches have been developed for recurrence prediction in previous studies, but most of them use only structured electronic health records and only a small training dataset, with limited success in clinical application. While free-text clinic notes may offer the greatest nuance and detail about a patient's clinical status, they are largely excluded in previous predictive models due to the increase in processing complexity and need for a complex modeling framework. In this study, we developed a weak-supervision framework for breast cancer recurrence prediction in which we trained a deep learning model on a large sample of free-text clinic notes by utilizing a combination of manually curated labels and NLP-generated non-perfect recurrence labels. The model was trained jointly on manually curated data from 670 patients and NLP-curated data of 8062 patients. It was validated on manually annotated data from 224 patients with recurrence and achieved 0.94 AUROC. This weak supervision approach allowed us to learn from a larger dataset using imperfect labels and ultimately provided greater accuracy compared to a smaller hand-curated dataset, with less manual effort invested in curation.

Sanyal Josh, Tariq Amara, Kurian Allison W, Rubin Daniel, Banerjee Imon

2021-May-04

Surgery Surgery

2-step deep learning model for landmarks localization in spine radiographs.

In Scientific reports ; h5-index 158.0

In this work we propose to use Deep Learning to automatically calculate the coordinates of the vertebral corners in sagittal x-rays images of the thoracolumbar spine and, from those landmarks, to calculate relevant radiological parameters such as L1-L5 and L1-S1 lordosis and sacral slope. For this purpose, we used 10,193 images annotated with the landmarks coordinates as the ground truth. We realized a model that consists of 2 steps. In step 1, we trained 2 Convolutional Neural Networks to identify each vertebra in the image and calculate the landmarks coordinates respectively. In step 2, we refined the localization using cropped images of a single vertebra as input to another convolutional neural network and we used geometrical transformations to map the corners to the original image. For the localization tasks, we used a differentiable spatial to numerical transform (DSNT) as the top layer. We evaluated the model both qualitatively and quantitatively on a set of 195 test images. The median localization errors relative to the vertebrae dimensions were 1.98% and 1.68% for x and y coordinates respectively. All the predicted angles were highly correlated with the ground truth, despite non-negligible absolute median errors of 1.84°, 2.43° and 1.98° for L1-L5, L1-S1 and SS respectively. Our model is able to calculate with good accuracy the coordinates of the vertebral corners and has a large potential for improving the reliability and repeatability of measurements in clinical tasks.

Cina Andrea, Bassani Tito, Panico Matteo, Luca Andrea, Masharawi Youssef, Brayda-Bruno Marco, Galbusera Fabio

2021-May-04

General General

Machine learning-based mortality prediction model for heat-related illness.

In Scientific reports ; h5-index 158.0

In this study, we aimed to develop and validate a machine learning-based mortality prediction model for hospitalized heat-related illness patients. After 2393 hospitalized patients were extracted from a multicentered heat-related illness registry in Japan, subjects were divided into the training set for development (n = 1516, data from 2014, 2017-2019) and the test set (n = 877, data from 2020) for validation. Twenty-four variables including characteristics of patients, vital signs, and laboratory test data at hospital arrival were trained as predictor features for machine learning. The outcome was death during hospital stay. In validation, the developed machine learning models (logistic regression, support vector machine, random forest, XGBoost) demonstrated favorable performance for outcome prediction with significantly increased values of the area under the precision-recall curve (AUPR) of 0.415 [95% confidence interval (CI) 0.336-0.494], 0.395 [CI 0.318-0.472], 0.426 [CI 0.346-0.506], and 0.528 [CI 0.442-0.614], respectively, compared to that of the conventional acute physiology and chronic health evaluation (APACHE)-II score of 0.287 [CI 0.222-0.351] as a reference standard. The area under the receiver operating characteristic curve (AUROC) values were also high over 0.92 in all models, although there were no statistical differences compared to APACHE-II. This is the first demonstration of the potential of machine learning-based mortality prediction models for heat-related illnesses.

Hirano Yohei, Kondo Yutaka, Hifumi Toru, Yokobori Shoji, Kanda Jun, Shimazaki Junya, Hayashida Kei, Moriya Takashi, Yagi Masaharu, Takauji Shuhei, Yamaguchi Junko, Okada Yohei, Okano Yuichi, Kaneko Hitoshi, Kobayashi Tatsuho, Fujita Motoki, Yokota Hiroyuki, Okamoto Ken, Tanaka Hiroshi, Yaguchi Arino

2021-May-04

General General

A data-driven approach to violin making.

In Scientific reports ; h5-index 158.0

Of all the characteristics of a violin, those that concern its shape are probably the most important ones, as the violin maker has complete control over them. Contemporary violin making, however, is still based more on tradition than understanding, and a definitive scientific study of the specific relations that exist between shape and vibrational properties is yet to come and sorely missed. In this article, using standard statistical learning tools, we show that the modal frequencies of violin tops can, in fact, be predicted from geometric parameters, and that artificial intelligence can be successfully applied to traditional violin making. We also study how modal frequencies vary with the thicknesses of the plate (a process often referred to as plate tuning) and discuss the complexity of this dependency. Finally, we propose a predictive tool for plate tuning, which takes into account material and geometric parameters.

Gonzalez Sebastian, Salvi Davide, Baeza Daniel, Antonacci Fabio, Sarti Augusto

2021-May-04

General General

Validation of Dairy Cow Bodyweight Prediction Using Traits Easily Recorded by Dairy Herd Improvement Organizations and Its Potential Improvement Using Feature Selection Algorithms.

In Animals : an open access journal from MDPI

Knowing the body weight (BW) of a cow at a specific moment or measuring its changes through time is of interest for management purposes. The current work aimed to validate the feasibility of predicting BW using the day in milk, parity, milk yield, and milk mid-infrared (MIR) spectrum from a multiple-country dataset and reduce the number of predictors to limit the risk of over-fitting and potentially improve its accuracy. The BW modeling procedure involved feature selections and herd-independent validation in identifying the most interesting subsets of predictors and then external validation of the models. From 1849 records collected in 9 herds from 360 Holstein cows, the best performing models achieved a root mean square error (RMSE) for the herd-independent validation between 52 ± 2.34 kg to 56 ± 3.16 kg, including from 5 to 62 predictors. Among these models, three performed remarkably well in external validation using an independent dataset (N = 4067), resulting in RMSE ranging from 52 to 56 kg. The results suggest that multiple optimal BW predictive models coexist due to the high correlations between adjacent spectral points.

Tedde Anthony, Grelet Clément, Ho Phuong N, Pryce Jennie E, Hailemariam Dagnachew, Wang Zhiquan, Plastow Graham, Gengler Nicolas, Brostaux Yves, Froidmont Eric, Dehareng Frédéric, Bertozzi Carlo, Crowe Mark A, Dufrasne Isabelle, Soyeurt Hélène

2021-Apr-30

dairy cow bodyweight, dairy cows, dimensionality reduction, feature selection, machine learning, mid infrared spectra, partial least square

General General

Detecting Topic and Sentiment Trends in Physician Rating Websites: Analysis of Online Reviews Using 3-Wave Datasets.

In International journal of environmental research and public health ; h5-index 73.0

(1) Background: Physician rating websites (PRWs) are a rich resource of information where individuals learn other people response to various health problems. The current study aims to investigate and analyze the people top concerns and sentiment dynamics expressed in physician online reviews (PORs). (2) Methods: Text data were collected from four U.S.-based PRWs during the three time periods of 2018, 2019 and 2020. Based on the dynamic topic modeling, hot topics related to different aspects of healthcare were identified. Following the hybrid approach of aspect-based sentiment analysis, the social network of prevailing topics was also analyzed whether people expressed positive, neutral or negative sentiments in PORs. (3) Results: The study identified 30 dominant topics across three different stages which lead toward four key findings. First, topics discussed in Stage III were quite different from the earlier two stages due to the COVID-19 outbreak. Second, based on the keyword co-occurrence analysis, the most prevalent keywords in all three stages were related to the treatment, questions asked by patients, communication problem, patients' feelings toward the hospital environment, disease symptoms, time spend with patients and different issues related to the COVID-19 (i.e., pneumonia, death, spread and cases). Third, topics related to the provider service quality, hospital servicescape and treatment cost were the most dominant topics in Stages I and II, while the quality of online information regarding COVID-19 and government countermeasures were the most dominant topics in Stage III. Fourth, when zooming into the topic-based sentiments analysis, hot topics in Stage I were mostly positive (joy be the dominant emotion), then negative (disgust be the dominant emotion) in Stage II. Furthermore, sentiments in the initial period of Stage III (COVID-19) were negative (anger be the dominant emotion), then transformed into positive (trust be the dominant emotion) later. The findings also revealed that the proposed method outperformed the conventional machine learning models in analyzing topic and sentiment dynamics expressed in PRWs. (4) Conclusions: Methodologically, this research demonstrates the ability and importance of computational techniques for analyzing large corpora of text and complementing conventional social science approaches.

Shah Adnan Muhammad, Naqvi Rizwan Ali, Jeong Ok-Ran

2021-Apr-29

COVID-19, discrete emotions, online reviews, sentiment analysis, text mining, topic modeling

General General

In the Danger Zone: U-Net Driven Quantile Regression can Predict High-risk SARS-CoV-2 Regions via Pollutant Particulate Matter and Satellite Imagery

ArXiv Preprint

Since the outbreak of COVID-19 policy makers have been relying upon non-pharmacological interventions to control the outbreak. With air pollution as a potential transmission vector there is need to include it in intervention strategies. We propose a U-net driven quantile regression model to predict $PM_{2.5}$ air pollution based on easily obtainable satellite imagery. We demonstrate that our approach can reconstruct $PM_{2.5}$ concentrations on ground-truth data and predict reasonable $PM_{2.5}$ values with their spatial distribution, even for locations where pollution data is unavailable. Such predictions of $PM_{2.5}$ characteristics could crucially advise public policy strategies geared to reduce the transmission of and lethality of COVID-19.

Jacquelyn Shelton, Przemyslaw Polewski, Wei Yao

2021-05-06

General General

In the Danger Zone: U-Net Driven Quantile Regression can Predict High-risk SARS-CoV-2 Regions via Pollutant Particulate Matter and Satellite Imagery

ArXiv Preprint

Since the outbreak of COVID-19 policy makers have been relying upon non-pharmacological interventions to control the outbreak. With air pollution as a potential transmission vector there is need to include it in intervention strategies. We propose a U-net driven quantile regression model to predict $PM_{2.5}$ air pollution based on easily obtainable satellite imagery. We demonstrate that our approach can reconstruct $PM_{2.5}$ concentrations on ground-truth data and predict reasonable $PM_{2.5}$ values with their spatial distribution, even for locations where pollution data is unavailable. Such predictions of $PM_{2.5}$ characteristics could crucially advise public policy strategies geared to reduce the transmission of and lethality of COVID-19.

Jacquelyn Shelton, Przemyslaw Polewski, Wei Yao

2021-05-06

General General

Machine learning based models for prediction of subtype diagnosis of primary aldosteronism using blood test.

In Scientific reports ; h5-index 158.0

Primary aldosteronism (PA) is associated with an increased risk of cardiometabolic diseases, especially in unilateral subtype. Despite its high prevalence, the case detection rate of PA is limited, partly because of no clinical models available in general practice to identify patients highly suspicious of unilateral subtype of PA, who should be referred to specialized centers. The aim of this retrospective cross-sectional study was to develop a predictive model for subtype diagnosis of PA based on machine learning methods using clinical data available in general practice. Overall, 91 patients with unilateral and 138 patients with bilateral PA were randomly assigned to the training and test cohorts. Four supervised machine learning classifiers; logistic regression, support vector machines, random forests (RF), and gradient boosting decision trees, were used to develop predictive models from 21 clinical variables. The accuracy and the area under the receiver operating characteristic curve (AUC) for predicting of subtype diagnosis of PA in the test cohort were compared among the optimized classifiers. Of the four classifiers, the accuracy and AUC were highest in RF, with 95.7% and 0.990, respectively. Serum potassium, plasma aldosterone, and serum sodium levels were highlighted as important variables in this model. For feature-selected RF with the three variables, the accuracy and AUC were 89.1% and 0.950, respectively. With an independent external PA cohort, we confirmed a similar accuracy for feature-selected RF (accuracy: 85.1%). Machine learning models developed using blood test can help predict subtype diagnosis of PA in general practice.

Kaneko Hiroki, Umakoshi Hironobu, Ogata Masatoshi, Wada Norio, Iwahashi Norifusa, Fukumoto Tazuru, Yokomoto-Umakoshi Maki, Nakano Yui, Matsuda Yayoi, Miyazawa Takashi, Sakamoto Ryuichi, Ogawa Yoshihiro

2021-May-04

oncology Oncology

Interrogation of gender disparity uncovers androgen receptor as the transcriptional activator for oncogenic miR-125b in gastric cancer.

In Cell death & disease

There is a male preponderance in gastric cancer (GC), which suggests a role of androgen and androgen receptor (AR). However, the mechanism of AR signaling in GC especially in female patients remains obscure. We sought to identify the AR signaling pathway that might be related to prognosis and examine the potential clinical utility of the AR antagonist for treatment. Deep learning and gene set enrichment analysis was used to identify potential critical factors associated with gender bias in GC (n = 1390). Gene expression profile analysis was performed to screen differentially expressed genes associated with AR expression in the Tianjin discovery set (n = 90) and TCGA validation set (n = 341). Predictors of survival were identified via lasso regression analyses and validated in the expanded Tianjin cohort (n = 373). In vitro and in vivo experiments were established to determine the drug effect. The GC gender bias was attributable to sex chromosome abnormalities and AR signaling dysregulation. The candidates for AR-related gene sets were screened, and AR combined with miR-125b was associated with poor prognosis, particularly among female patients. AR was confirmed to directly regulate miR-125b expression. AR-miR-125b signaling pathway inhibited apoptosis and promoted proliferation. AR antagonist, bicalutamide, exerted anti-tumor activities and induced apoptosis both in vitro and in vivo, using GC cell lines and female patient-derived xenograft (PDX) model. We have shed light on gender differences by revealing a hormone-regulated oncogenic signaling pathway in GC. Our preclinical studies suggest that AR is a potential therapeutic target for this deadly cancer type, especially in female patients.

Liu Ben, Zhou Meng, Li Xiangchun, Zhang Xining, Wang Qinghua, Liu Luyang, Yang Meng, Yang Da, Guo Yan, Zhang Qiang, Zheng Hong, Wang Qiong, Li Lian, Chu Xinlei, Wang Wei, Li Haixin, Song Fengju, Pan Yuan, Zhang Wei, Chen Kexin

2021-May-04

General General

Metagenomic Analysis of Common Intestinal Diseases Reveals Relationships among Microbial Signatures and Powers Multidisease Diagnostic Models.

In mSystems

Common intestinal diseases such as Crohn's disease (CD), ulcerative colitis (UC), and colorectal cancer (CRC) share clinical symptoms and altered gut microbes, necessitating cross-disease comparisons and the use of multidisease models. Here, we performed meta-analyses on 13 fecal metagenome data sets of the three diseases. We identified 87 species and 65 pathway markers that were consistently changed in multiple data sets of the same diseases. According to their overall trends, we grouped the disease-enriched marker species into disease-specific and disease-common clusters and revealed their distinct phylogenetic relationships; species in the CD-specific cluster were phylogenetically related, while those in the CRC-specific cluster were more distant. Strikingly, UC-specific species were phylogenetically closer to CRC, likely because UC patients have higher risk of CRC. Consistent with their phylogenetic relationships, marker species had similar within-cluster and different between-cluster metabolic preferences. A portion of marker species and pathways correlated with an indicator of leaky gut, suggesting a link between gut dysbiosis and human-derived contents. Marker species showed more coordinated changes and tighter inner-connections in cases than the controls, suggesting that the diseased gut may represent a stressed environment and pose stronger selection on gut microbes. With the marker species and pathways, we constructed four high-performance (including multidisease) models with an area under the receiver operating characteristic curve (AUROC) of 0.87 and true-positive rates up to 90%, and explained their putative clinical applications. We identified consistent microbial alterations in common intestinal diseases, revealed metabolic capacities and the relationships among marker bacteria in distinct states, and supported the feasibility of metagenome-derived multidisease diagnosis.IMPORTANCE Gut microbes have been identified as potential markers in distinguishing patients from controls in colorectal cancer, ulcerative colitis, and Crohn's disease individually, whereas there lacks a systematic analysis to investigate the exclusive microbial shifts of these enteropathies with similar clinical symptoms. Our meta-analysis and cross-disease comparisons identified consistent microbial alterations in each enteropathy, revealed microbial ecosystems among marker bacteria in distinct states, and demonstrated the necessity and feasibility of metagenome-based multidisease classifications. To the best of our knowledge, this is the first study to construct multiclass models for these common intestinal diseases.

Jiang Puzi, Wu Sicheng, Luo Qibin, Zhao Xing-Ming, Chen Wei-Hua

2021-May-04

gut dysbiosis, human microbiome, intestinal disease, machine learning-based disease classification, noninvasive disease diagnosis

oncology Oncology

Deep learning for fully-automated prediction of overall survival in patients with oropharyngeal cancer using FDG PET imaging: an international retrospective study.

In Clinical cancer research : an official journal of the American Association for Cancer Research

PURPOSE : Accurate prognostic stratification of patients with oropharyngeal squamous cell carcinoma (OPSCC) is crucial. We developed an objective and robust deep learning-based fully-automated tool called the DeepPET-OPSCC biomarker for predicting overall survival (OS) in OPSCC using [18F]fluorodeoxyglucose PET imaging.

EXPERIMENTAL DESIGN : The DeepPET-OPSCC prediction model was built and tested internally on a discovery cohort (n = 268) by integrating five convolutional neural network models for volumetric segmentation and ten models for OS prognostication. Two external test cohorts were enrolled - the first based on the Cancer Imaging Archive (TCIA) database (n = 353) and the second being a clinical deployment cohort (n = 31) - to assess the DeepPET-OPSCC performance and goodness of fit.

RESULTS : After adjustment for potential confounders, DeepPET-OPSCC was found to be an independent predictor of OS in both discovery and TCIA test cohorts (HR = 2.07; 95% CI 1.31-3.28 and HR = 2.39; 1.38-4.16; both P = 0.002). The tool also revealed good predictive performance, with a c-index of 0.707 (95% CI 0.658-0.757) in the discovery cohort, 0.689 (0.621-0.757) in the TCIA test cohort, and 0.787 (0.675-0.899) in the clinical deployment test cohort; the average time taken was 2 min for calculation per exam. The integrated nomogram of DeepPET-OPSCC and clinical risk factors significantly outperformed the clinical model (AUC at 5 years: 0.801 [95% CI 0.727-0.874] versus 0.749 [0.649-0.842]; P = 0.031) in the TCIA test cohort.

CONCLUSIONS : DeepPET-OPSCC achieved an accurate OS prediction in patients with OPSCC and enabled an objective, unbiased, and rapid assessment for OPSCC prognostication.

Cheng Nai-Ming, Yao Jiawen, Cai Jinzheng, Ye Xianghua, Zhao Shilin, Zhao Kui, Zhou Wenlan, Nogues Isabella, Huo Yuankai, Liao Chun-Ta, Wang Hung-Ming, Lin Chien-Yu, Lee Li-Yu, Xiao Jing, Lu Le, Zhang Ling, Yen Tzu-Chen

2021-May-04

General General

Natural statistics as inference principles of auditory tuning in biological and artificial midbrain networks.

In eNeuro

Bats provide a powerful mammalian model to explore the neural representation of complex sounds, as they rely on hearing to survive in their environment. The inferior colliculus (IC) is a central hub of the auditory system that receives converging projections from the ascending pathway and descending inputs from auditory cortex. In this work, we build an artificial neural network to replicate auditory characteristics in IC neurons of the big brown bat. We first test the hypothesis that spectro-temporal tuning of IC neurons is optimized to represent the natural statistics of conspecific vocalizations. We estimate spectro-temporal receptive fields (STRF) of IC neurons and compare tuning characteristics to statistics of bat calls. The results indicate that the FM tuning of IC neurons is matched with the statistics. Then, we investigate this hypothesis on the network optimized to represent natural sound statistics and to compare its output with biological responses. We also estimate biomimetic STRF's from the artificial network and correlate their characteristics to those of biological neurons. Tuning properties of both biological and artificial neurons reveal strong agreement along both spectral and temporal dimensions, and suggest the presence of nonlinearity, sparsity and complexity constraints that underlie the neural representation in the auditory midbrain. Additionally, the artificial neurons replicate IC neural activities in discrimination of social calls, and provide simulated results for a noise robust discrimination. In this way, the biomimetic network allows us to infer the neural mechanisms by which the bat's IC processes natural sounds used to construct the auditory scene.Significance StatementRecent advances in machine learning have led to powerful mathematical mappings of complex data. Applied to brain structures, artificial neural networks can be configured to explore principles underlying neural encoding of complex stimuli. Bats use a rich repertoire of calls to communicate and navigate their world, and the statistics underlying the calls appear to align with tuning selectivity of neurons. We show that artificial neural network with a nonlinear, sparse and deep architecture trained on the statistics of bat communication and echolocation calls results in a close match to neurons from bat's inferior colliculus. This tuning optimized to yield an effective representation of spectro-temporal statistics of bat calls appears to underlie strong selectivity and noise invariance in the inferior colliculus.

Park Sangwook, Salles Angeles, Allen Kathryne, Moss Cynthia F, Elhilali Mounya

2021-May-04

Big brown bat, Biomimetic Network, IC, Spectrotemporal receptive fields, machine learning

Surgery Surgery

Improving burn depth assessment for pediatric scalds by AI based on semantic segmentation of polarized light photography images.

In Burns : journal of the International Society for Burn Injuries

This paper illustrates the efficacy of an artificial intelligence (AI) (a convolutional neural network, based on the U-Net), for the burn-depth assessment using semantic segmentation of polarized high-performance light camera images of burn wounds. The proposed method is evaluated for paediatric scald injuries to differentiate four burn wound depths: superficial partial-thickness (healing in 0-7 days), superficial to intermediate partial-thickness (healing in 8-13 days), intermediate to deep partial-thickness (healing in 14-20 days), deep partial-thickness (healing after 21 days) and full-thickness burns, based on observed healing time. In total 100 burn images were acquired. Seventeen images contained all 4 burn depths and were used to train the network. Leave-one-out cross-validation reports were generated and an accuracy and dice coefficient average of almost 97% was then obtained. After that, the remaining 83 burn-wound images were evaluated using the different network during the cross-validation, achieving an accuracy and dice coefficient, both on average 92%. This technique offers an interesting new automated alternative for clinical decision support to assess and localize burn-depths in 2D digital images. Further training and improvement of the underlying algorithm by e.g., more images, seems feasible and thus promising for the future.

Cirillo Marco Domenico, Mirdell Robin, Sjöberg Folke, Pham Tuan D

2021-Feb-08

Artificial intelligence, Convolutional neural networks, Deep learning, Paediatric burns, Semantic segmentation, U-Net

General General

Neural correlates of treatment effect and prediction of treatment outcome in patients with PTSD and comorbid personality disorder: study design.

In Borderline personality disorder and emotion dysregulation

BACKGROUND : Neural alterations related to treatment outcome in patients with both post-traumatic stress disorder (PTSD) and comorbid personality disorder are unknown. Here we describe the protocol for a neuroimaging study of treatment of patients with PTSD and comorbid borderline (BPD) or cluster C (CPD) personality disorder traits. Our specific aims are to 1) investigate treatment-induced neural alterations, 2) predict treatment outcome using structural and functional magnetic resonance imaging (MRI) and 3) study neural alterations associated with BPD and CPD in PTSD patients. We hypothesize that 1) all treatment conditions are associated with normalization of limbic and prefrontal brain activity and hyperconnectivity in resting-state brain networks, with additional normalization of task-related activation in emotion regulation brain areas in the patients who receive trauma-focused therapy and personality disorder treatment; 2) Baseline task-related activation, together with structural brain measures and clinical variables predict treatment outcome; 3) dysfunction in task-related activation and resting-state connectivity of emotion regulation areas is comparable in PTSD patients with BPD or CPD, with a hypoconnected central executive network in patients with PTSD+BPD.

METHODS : We aim to include pre- and post-treatment 3 T-MRI scans in 40 patients with PTSD and (sub) clinical comorbid BPD or CPD. With an expected attrition rate of 50%, at least 80 patients will be scanned before treatment. MRI scans for 30 matched healthy controls will additionally be acquired. Patients with PTSD and BPD were randomized to either EMDR-only or EMDR combined with Dialectical Behaviour Therapy. Patients with PTSD and CPD were randomized to Imaginary Rescripting (ImRs) or to ImRs combined with Schema Focused Therapy. The scan protocol consists of a T1-weighted structural scan, resting state fMRI, task-based fMRI during an emotional face task and multi-shell diffusion weighted images. For data analysis, multivariate mixed-models, regression analyses and machine learning models will be used.

DISCUSSION : This study is one of the first to use neuroimaging measures to predict and better understand treatment response in patients with PTSD and comorbid personality disorders. A heterogeneous, naturalistic sample will be included, ensuring generalizability to a broad group of treatment seeking PTSD patients.

TRIAL REGISTRATION : Clinical Trials, NCT03833453 & NCT03833531 . Retrospectively registered, February 2019.

Aarts Inga, Vriend Chris, Snoek Aishah, van den End Arne, Blankers Matthijs, Beekman Aartjan T F, Dekker Jack, van den Heuvel Odile A, Thomaes Kathleen

2021-May-05

Borderline personality disorder, Cluster C personality disorder, Neuroimaging, Prediction, Treatment

Cardiology Cardiology

Glycemic and lipid variability for predicting complications and mortality in diabetes mellitus using machine learning.

In BMC endocrine disorders

INTRODUCTION : Recent studies have reported that HbA1c and lipid variability is useful for risk stratification in diabetes mellitus. The present study evaluated the predictive value of the baseline, subsequent mean of at least three measurements and variability of HbA1c and lipids for adverse outcomes.

METHODS : This retrospective cohort study consists of type 1 and type 2 diabetic patients who were prescribed insulin at outpatient clinics of Hong Kong public hospitals, from 1st January to 31st December 2009. Standard deviation (SD) and coefficient of variation were used to measure the variability of HbA1c, total cholesterol, low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C) and triglyceride. The primary outcome is all-cause mortality. Secondary outcomes were diabetes-related complications.

RESULT : The study consists of 25,186 patients (mean age = 63.0, interquartile range [IQR] of age = 15.1 years, male = 50%). HbA1c and lipid value and variability were significant predictors of all-cause mortality. Higher HbA1c and lipid variability measures were associated with increased risks of neurological, ophthalmological and renal complications, as well as incident dementia, osteoporosis, peripheral vascular disease, ischemic heart disease, atrial fibrillation and heart failure (p <  0.05). Significant association was found between hypoglycemic frequency (p <  0.0001), HbA1c (p <  0.0001) and lipid variability against baseline neutrophil-lymphocyte ratio (NLR).

CONCLUSION : Raised variability in HbA1c and lipid parameters are associated with an elevated risk in both diabetic complications and all-cause mortality. The association between hypoglycemic frequency, baseline NLR, and both HbA1c and lipid variability implicate a role for inflammation in mediating adverse outcomes in diabetes, but this should be explored further in future studies.

Lee Sharen, Zhou Jiandong, Wong Wing Tak, Liu Tong, Wu William K K, Wong Ian Chi Kei, Zhang Qingpeng, Tse Gary

2021-May-04

General General

A filter approach for feature selection in classification: application to automatic atrial fibrillation detection in electrocardiogram recordings.

In BMC medical informatics and decision making ; h5-index 38.0

BACKGROUND : In high-dimensional data analysis, the complexity of predictive models can be reduced by selecting the most relevant features, which is crucial to reduce data noise and increase model accuracy and interpretability. Thus, in the field of clinical decision making, only the most relevant features from a set of medical descriptors should be considered when determining whether a patient is healthy or not. This statistical approach known as feature selection can be performed through regression or classification, in a supervised or unsupervised manner. Several feature selection approaches using different mathematical concepts have been described in the literature. In the field of classification, a new approach has recently been proposed that uses the [Formula: see text]-metric, an index measuring separability between different classes in heart rhythm characterization. The present study proposes a filter approach for feature selection in classification using this [Formula: see text]-metric, and evaluates its application to automatic atrial fibrillation detection.

METHODS : The stability and prediction performance of the [Formula: see text]-metric feature selection approach was evaluated using the support vector machine model on two heart rhythm datasets, one extracted from the PhysioNet database and the other from the database of Marseille University Hospital Center, France (Timone Hospital). Both datasets contained electrocardiogram recordings grouped into two classes: normal sinus rhythm and atrial fibrillation. The performance of this feature selection approach was compared to that of three other approaches, with the first two based on the Random Forest technique and the other on receiver operating characteristic curve analysis.

RESULTS : The [Formula: see text]-metric approach showed satisfactory results, especially for models with a smaller number of features. For the training dataset, all prediction indicators were higher for our approach (accuracy greater than 99% for models with 5 to 17 features), as was stability (greater than 0.925 regardless of the number of features included in the model). For the validation dataset, the features selected with the [Formula: see text]-metric approach differed from those selected with the other approaches; sensitivity was higher for our approach, but other indicators were similar.

CONCLUSION : This filter approach for feature selection in classification opens up new methodological avenues for atrial fibrillation detection using short electrocardiogram recordings.

Michel Pierre, Ngo Nicolas, Pons Jean-François, Delliaux Stéphane, Giorgi Roch

2021-May-04

-metric, Atrial fibrillation detection, Classification, Clinical decision making, Feature selection, Machine learning

General General

Decision making under ambiguity and risk in adolescent-onset schizophrenia.

In BMC psychiatry

OBJECTIVE : Numerous studies have identified impaired decision making (DM) under both ambiguity and risk in adult patients with schizophrenia. However, the assessment of DM in patients with adolescent-onset schizophrenia (AOS) has been challenging as a result of the instability and heterogeneity of manifestations. The Iowa Gambling Task (IGT) and Game of Dice Task (GDT), which are frequently used to evaluate DM respectively under ambiguity and risk, are sensitive to adolescents and neuropsychiatric patients. Our research intended to examine the performance of DM in a relatively large sample of patients with AOS using the above-mentioned two tasks. We also aimed to take a closer look at the relationship between DM and symptom severity of schizophrenia.

METHODS : We compared the performance of DM in 71 patients with AOS and 53 well-matched healthy controls using IGT for DM under ambiguity and GDT for DM under risk through net scores, total scores and feedback ration. Neuropsychological tests were conducted in all participants. Clinical symptoms were evaluated by using Positive and Negative Syndrome Scale (PANSS) in 71 patients with AOS. Pearson's correlation revealed the relationship among total score of DM and clinical and neuropsychological data.

RESULTS : Compared to healthy controls, patients with AOS failed to show learning effect and had a significant difference on the 5th block in IGT and conducted more disadvantageous choices as well as exhibited worse negative feedback rate in GDT. Apart from DM impairment under risk, diminished DM abilities under ambiguity were found related to poor executive function in AOS in the present study.

CONCLUSIONS : Our findings unveiled the abnormal pattern of DM in AOS, mainly reflected under the risky condition, extending the knowledge on the performance of DM under ambiguity and risk in AOS. Inefficient DM under risk may account for the lagging impulse control and the combined effects of developmental disease. In addition, our study demonstrated that the performance on IGT was related to executive function in AOS.

Li Dandan, Zhang Fengyan, Wang Lu, Zhang Yifan, Yang Tingting, Wang Kai, Zhu Chunyan

2021-05-04

Adolescent-onset schizophrenia, Decision making, Game of dice task, Iowa gambling task

Radiology Radiology

An Overview of Artificial Intelligence Applications in Liver and Pancreatic Imaging.

In Cancers

Artificial intelligence (AI) is one of the most promising fields of research in medical imaging so far. By means of specific algorithms, it can be used to help radiologists in their routine workflow. There are several papers that describe AI approaches to solve different problems in liver and pancreatic imaging. These problems may be summarized in four different categories: segmentation, quantification, characterization and image quality improvement. Segmentation is usually the first step of successive elaborations. If done manually, it is a time-consuming process. Therefore, the semi-automatic and automatic creation of a liver or a pancreatic mask may save time for other evaluations, such as quantification of various parameters, from organs volume to their textural features. The alterations of normal liver and pancreas structure may give a clue to the presence of a diffuse or focal pathology. AI can be trained to recognize these alterations and propose a diagnosis, which may then be confirmed or not by radiologists. Finally, AI may be applied in medical image reconstruction in order to increase image quality, decrease dose administration (referring to computed tomography) and reduce scan times. In this article, we report the state of the art of AI applications in these four main categories.

Cardobi Nicolò, Dal Palù Alessandro, Pedrini Federica, Beleù Alessandro, Nocini Riccardo, De Robertis Riccardo, Ruzzenente Andrea, Salvia Roberto, Montemezzi Stefania, D’Onofrio Mirko

2021-Apr-30

artificial intelligence, deep learning, liver imaging, machine learning, pancreatic imaging

Radiology Radiology

A Novel Method for COVID-19 Diagnosis Using Artificial Intelligence in Chest X-ray Images.

In Healthcare (Basel, Switzerland)

The Coronavirus disease 2019 (COVID-19) is an infectious disease spreading rapidly and uncontrollably throughout the world. The critical challenge is the rapid detection of Coronavirus infected people. The available techniques being utilized are body-temperature measurement, along with anterior nasal swab analysis. However, taking nasal swabs and lab testing are complex, intrusive, and require many resources. Furthermore, the lack of test kits to meet the exceeding cases is also a major limitation. The current challenge is to develop some technology to non-intrusively detect the suspected Coronavirus patients through Artificial Intelligence (AI) techniques such as deep learning (DL). Another challenge to conduct the research on this area is the difficulty of obtaining the dataset due to a limited number of patients giving their consent to participate in the research study. Looking at the efficacy of AI in healthcare systems, it is a great challenge for the researchers to develop an AI algorithm that can help health professionals and government officials automatically identify and isolate people with Coronavirus symptoms. Hence, this paper proposes a novel method CoVIRNet (COVID Inception-ResNet model), which utilizes the chest X-rays to diagnose the COVID-19 patients automatically. The proposed algorithm has different inception residual blocks that cater to information by using different depths feature maps at different scales, with the various layers. The features are concatenated at each proposed classification block, using the average-pooling layer, and concatenated features are passed to the fully connected layer. The efficient proposed deep-learning blocks used different regularization techniques to minimize the overfitting due to the small COVID-19 dataset. The multiscale features are extracted at different levels of the proposed deep-learning model and then embedded into various machine-learning models to validate the combination of deep-learning and machine-learning models. The proposed CoVIR-Net model achieved 95.7% accuracy, and the CoVIR-Net feature extractor with random-forest classifier produced 97.29% accuracy, which is the highest, as compared to existing state-of-the-art deep-learning methods. The proposed model would be an automatic solution for the assessment and classification of COVID-19. We predict that the proposed method will demonstrate an outstanding performance as compared to the state-of-the-art techniques being used currently.

Almalki Yassir Edrees, Qayyum Abdul, Irfan Muhammad, Haider Noman, Glowacz Adam, Alshehri Fahad Mohammed, Alduraibi Sharifa K, Alshamrani Khalaf, Alkhalik Basha Mohammad Abd, Alduraibi Alaa, Saeed M K, Rahman Saifur

2021-Apr-29

chest X-ray images, data analytics, feature extraction, healthcare, image processing, pandemic

Radiology Radiology

Quantification of pulmonary involvement in COVID-19 pneumonia by means of a cascade oftwo U-nets: training and assessment on multipledatasets using different annotation criteria

ArXiv Preprint

The automatic assignment of a severity score to the CT scans of patients affected by COVID-19 pneumonia could reduce the workload in radiology departments. This study aims at exploiting Artificial intelligence (AI) for the identification, segmentation and quantification of COVID-19 pulmonary lesions. We investigated the effects of using multiple datasets, heterogeneously populated and annotated according to different criteria. We developed an automated analysis pipeline, the LungQuant system, based on a cascade of two U-nets. The first one (U-net_1) is devoted to the identification of the lung parenchyma, the second one (U-net_2) acts on a bounding box enclosing the segmented lungs to identify the areas affected by COVID-19 lesions. Different public datasets were used to train the U-nets and to evaluate their segmentation performances, which have been quantified in terms of the Dice index. The accuracy in predicting the CT-Severity Score (CT-SS) of the LungQuant system has been also evaluated. Both Dice and accuracy showed a dependency on the quality of annotations of the available data samples. On an independent and publicly available benchmark dataset, the Dice values measured between the masks predicted by LungQuant system and the reference ones were 0.95$\pm$0.01 and 0.66$\pm$0.13 for the segmentation of lungs and COVID-19 lesions, respectively. The accuracy of 90% in the identification of the CT-SS on this benchmark dataset was achieved. We analysed the impact of using data samples with different annotation criteria in training an AI-based quantification system for pulmonary involvement in COVID-19 pneumonia. In terms of the Dice index, the U-net segmentation quality strongly depends on the quality of the lesion annotations. Nevertheless, the CT-SS can be accurately predicted on independent validation sets, demonstrating the satisfactory generalization ability of the LungQuant.

Francesca Lizzi, Abramo Agosti, Francesca Brero, Raffaella Fiamma Cabini, Maria Evelina Fantacci, Silvia Figini, Alessandro Lascialfari, Francesco Laruina, Piernicola Oliva, Stefano Piffer, Ian Postuma, Lisa Rinaldi, Cinzia Talamonti, Alessandra Retico

2021-05-06

Radiology Radiology

Quantification of pulmonary involvement in COVID-19 pneumonia by means of a cascade oftwo U-nets: training and assessment on multipledatasets using different annotation criteria

ArXiv Preprint

The automatic assignment of a severity score to the CT scans of patients affected by COVID-19 pneumonia could reduce the workload in radiology departments. This study aims at exploiting Artificial intelligence (AI) for the identification, segmentation and quantification of COVID-19 pulmonary lesions. We investigated the effects of using multiple datasets, heterogeneously populated and annotated according to different criteria. We developed an automated analysis pipeline, the LungQuant system, based on a cascade of two U-nets. The first one (U-net_1) is devoted to the identification of the lung parenchyma, the second one (U-net_2) acts on a bounding box enclosing the segmented lungs to identify the areas affected by COVID-19 lesions. Different public datasets were used to train the U-nets and to evaluate their segmentation performances, which have been quantified in terms of the Dice index. The accuracy in predicting the CT-Severity Score (CT-SS) of the LungQuant system has been also evaluated. Both Dice and accuracy showed a dependency on the quality of annotations of the available data samples. On an independent and publicly available benchmark dataset, the Dice values measured between the masks predicted by LungQuant system and the reference ones were 0.95$\pm$0.01 and 0.66$\pm$0.13 for the segmentation of lungs and COVID-19 lesions, respectively. The accuracy of 90% in the identification of the CT-SS on this benchmark dataset was achieved. We analysed the impact of using data samples with different annotation criteria in training an AI-based quantification system for pulmonary involvement in COVID-19 pneumonia. In terms of the Dice index, the U-net segmentation quality strongly depends on the quality of the lesion annotations. Nevertheless, the CT-SS can be accurately predicted on independent validation sets, demonstrating the satisfactory generalization ability of the LungQuant.

Francesca Lizzi, Abramo Agosti, Francesca Brero, Raffaella Fiamma Cabini, Maria Evelina Fantacci, Silvia Figini, Alessandro Lascialfari, Francesco Laruina, Piernicola Oliva, Stefano Piffer, Ian Postuma, Lisa Rinaldi, Cinzia Talamonti, Alessandra Retico

2021-05-06

General General

Animal immunization merges with innovative technologies: A new paradigm shift in antibody discovery.

In mAbs

Animal-derived antibody sources, particularly, transgenic mice that are engineered with human immunoglobulin loci, along with advanced antibody generation technology platforms have facilitated the discoveries of human antibody therapeutics. For example, isolation of antigen-specific B cells, microfluidics, and next-generation sequencing have emerged as powerful tools for identifying and developing monoclonal antibodies (mAbs). These technologies enable not only antibody drug discovery but also lead to the understanding of B cell biology, immune mechanisms and immunogenetics of antibodies. In this perspective article, we discuss the scientific merits of animal immunization combined with advanced methods for antibody generation as compared to animal-free alternatives through in-vitro-generated antibody libraries. The knowledge gained from animal-derived antibodies concerning the recombinational diversity, somatic hypermutation patterns, and physiochemical properties is found more valuable and prerequisite for developing in vitro libraries, as well as artificial intelligence/machine learning methods to discover safe and effective mAbs.

Prabakaran Ponraj, Rao Sambasiva P, Wendt Maria

Animal immunization, animal-derived antibody, antibody discovery, antibody technologies, monoclonal antibody (mAb)

General General

Skin Lesion Segmentation and Multiclass Classification Using Deep Learning Features and Improved Moth Flame Optimization.

In Diagnostics (Basel, Switzerland)

Manual diagnosis of skin cancer is time-consuming and expensive; therefore, it is essential to develop automated diagnostics methods with the ability to classify multiclass skin lesions with greater accuracy. We propose a fully automated approach for multiclass skin lesion segmentation and classification by using the most discriminant deep features. First, the input images are initially enhanced using local color-controlled histogram intensity values (LCcHIV). Next, saliency is estimated using a novel Deep Saliency Segmentation method, which uses a custom convolutional neural network (CNN) of ten layers. The generated heat map is converted into a binary image using a thresholding function. Next, the segmented color lesion images are used for feature extraction by a deep pre-trained CNN model. To avoid the curse of dimensionality, we implement an improved moth flame optimization (IMFO) algorithm to select the most discriminant features. The resultant features are fused using a multiset maximum correlation analysis (MMCA) and classified using the Kernel Extreme Learning Machine (KELM) classifier. The segmentation performance of the proposed methodology is analyzed on ISBI 2016, ISBI 2017, ISIC 2018, and PH2 datasets, achieving an accuracy of 95.38%, 95.79%, 92.69%, and 98.70%, respectively. The classification performance is evaluated on the HAM10000 dataset and achieved an accuracy of 90.67%. To prove the effectiveness of the proposed methods, we present a comparison with the state-of-the-art techniques.

Khan Muhammad Attique, Sharif Muhammad, Akram Tallha, Damaševičius Robertas, Maskeliūnas Rytis

2021-Apr-29

deep features, feature fusion, heuristic feature optimization, melanoma, moth flame optimization, skin cancer

General General

Design of Nonlinear Autoregressive Exogenous Model Based Intelligence Computing for Efficient State Estimation of Underwater Passive Target.

In Entropy (Basel, Switzerland)

In this study, an intelligent computing paradigm built on a nonlinear autoregressive exogenous (NARX) feedback neural network model with the strength of deep learning is presented for accurate state estimation of an underwater passive target. In underwater scenarios, real-time motion parameters of passive objects are usually extracted with nonlinear filtering techniques. In filtering algorithms, nonlinear passive measurements are associated with linear kinetics of the target, governing by state space methodology. To improve tracking accuracy, effective feature estimation and minimizing position error of dynamic passive objects, the strength of NARX based supervised learning is exploited. Dynamic artificial neural networks, which contain tapped delay lines, are suitable for predicting the future state of the underwater passive object. Neural networks-based intelligence computing is effectively applied for estimating the real-time actual state of a passive moving object, which follows a semi-curved path. Performance analysis of NARX based neural networks is evaluated for six different scenarios of standard deviation of white Gaussian measurement noise by following bearings only tracking phenomena. Root mean square error between estimated and real position of the passive target in rectangular coordinates is computed for evaluating the worth of the proposed NARX feedback neural network scheme. The Monte Carlo simulations are conducted and the results certify the capability of the intelligence computing over conventional nonlinear filtering algorithms such as spherical radial cubature Kalman filter and unscented Kalman filter for given state estimation model.

Ali Wasiq, Khan Wasim Ullah, Raja Muhammad Asif Zahoor, He Yigang, Li Yaan

2021-Apr-29

artificial neural network, intelligent computing, measurement noise, nonlinear autoregressive with exogenous input (NARX), nonlinear filtering, state estimation

Cardiology Cardiology

Artificial Intelligence Applications for COVID-19 in Intensive Care and Emergency Settings: A Systematic Review.

In International journal of environmental research and public health ; h5-index 73.0

Background: Little is known about the role of artificial intelligence (AI) as a decisive technology in the clinical management of COVID-19 patients. We aimed to systematically review and critically appraise the current evidence on AI applications for COVID-19 in intensive care and emergency settings. Methods: We systematically searched PubMed, Embase, Scopus, CINAHL, IEEE Xplore, and ACM Digital Library databases from inception to 1 October 2020, without language restrictions. We included peer-reviewed original studies that applied AI for COVID-19 patients, healthcare workers, or health systems in intensive care, emergency, or prehospital settings. We assessed predictive modelling studies and critically appraised the methodology and key findings of all other studies. Results: Of fourteen eligible studies, eleven developed prognostic or diagnostic AI predictive models, all of which were assessed to be at high risk of bias. Common pitfalls included inadequate sample sizes, poor handling of missing data, failure to account for censored participants, and weak validation of models. Conclusions: Current AI applications for COVID-19 are not ready for deployment in acute care settings, given their limited scope and poor quality. Our findings underscore the need for improvements to facilitate safe and effective clinical adoption of AI applications, for and beyond the COVID-19 pandemic.

Chee Marcel Lucas, Ong Marcus Eng Hock, Siddiqui Fahad Javaid, Zhang Zhongheng, Lim Shir Lynn, Ho Andrew Fu Wah, Liu Nan

2021-Apr-29

COVID-19, artificial intelligence, critical care, emergency department, intensive care, machine learning

General General

Comparison of Targeted and Untargeted Approaches in Breath Analysis for the Discrimination of Lung Cancer from Benign Pulmonary Diseases and Healthy Persons.

In Molecules (Basel, Switzerland)

The aim of the present study was to compare the efficiency of targeted and untargeted breath analysis in the discrimination of lung cancer (Ca+) patients from healthy people (HC) and patients with benign pulmonary diseases (Ca-). Exhaled breath samples from 49 Ca+ patients, 36 Ca- patients and 52 healthy controls (HC) were analyzed by an SPME-GC-MS method. Untargeted treatment of the acquired data was performed with the use of the web-based platform XCMS Online combined with manual reprocessing of raw chromatographic data. Machine learning methods were applied to estimate the efficiency of breath analysis in the classification of the participants. Results: Untargeted analysis revealed 29 informative VOCs, from which 17 were identified by mass spectra and retention time/retention index evaluation. The untargeted analysis yielded slightly better results in discriminating Ca+ patients from HC (accuracy: 91.0%, AUC: 0.96 and accuracy 89.1%, AUC: 0.97 for untargeted and targeted analysis, respectively) but significantly improved the efficiency of discrimination between Ca+ and Ca- patients, increasing the accuracy of the classification from 52.9 to 75.3% and the AUC from 0.55 to 0.82. Conclusions: The untargeted breath analysis through the inclusion and utilization of newly identified compounds that were not considered in targeted analysis allowed the discrimination of the Ca+ from Ca- patients, which was not achieved by the targeted approach.

Koureas Michalis, Kalompatsios Dimitrios, Amoutzias Grigoris D, Hadjichristodoulou Christos, Gourgoulianis Konstantinos, Tsakalof Andreas

2021-Apr-29

breath analysis, cancer biomarkers, exhaled breath, lung cancer, untargeted analysis, volatile organic compounds, volatolomics

Surgery Surgery

Influence of the Depth of the Convolutional Neural Networks on an Artificial Intelligence Model for Diagnosis of Orthognathic Surgery.

In Journal of personalized medicine

The aim of this study was to investigate the relationship between image patterns in cephalometric radiographs and the diagnosis of orthognathic surgery and propose a method to improve the accuracy of predictive models according to the depth of the neural networks. The study included 640 and 320 patients requiring non-surgical and surgical orthodontic treatments, respectively. The data of 150 patients were exclusively classified as a test set. The data of the remaining 810 patients were split into five groups and a five-fold cross-validation was performed. The convolutional neural network models used were ResNet-18, 34, 50, and 101. The number in the model name represents the difference in the depth of the blocks that constitute the model. The accuracy, sensitivity, and specificity of each model were estimated and compared. The average success rate in the test set for the ResNet-18, 34, 50, and 101 was 93.80%, 93.60%, 91.13%, and 91.33%, respectively. In screening, ResNet-18 had the best performance with an area under the curve of 0.979, followed by ResNets-34, 50, and 101 at 0.974, 0.945, and 0.944, respectively. This study suggests the required characteristics of the structure of an artificial intelligence model for decision-making based on medical images.

Kim Ye-Hyun, Park Jae-Bong, Chang Min-Seok, Ryu Jae-Jun, Lim Won Hee, Jung Seok-Ki

2021-Apr-29

artificial intelligence, cephalometric analysis, convolutional neural network, deep learning, orthognathic surgery diagnosis

Radiology Radiology

Using Genomics Feature Selection Method in Radiomics Pipeline Improves Prognostication Performance in Locally Advanced Esophageal Squamous Cell Carcinoma-A Pilot Study.

In Cancers

PURPOSE : To evaluate the prognostic value of baseline and restaging CT-based radiomics with features associated with gene expression in esophageal squamous cell carcinoma (ESCC) patients receiving neoadjuvant chemoradiation (nCRT) plus surgery.

METHODS : We enrolled 106 ESCC patients receiving nCRT from two institutions. Gene expression profiles of 28 patients in the training set were used to detect differentially expressed (DE) genes between patients with and without relapse. Radiomic features that were correlated to DE genes were selected, followed by additional machine learning selection. A radiomic nomogram for disease-free survival (DFS) prediction incorporating the radiomic signature and prognostic clinical characteristics was established for DFS estimation and validated.

RESULTS : The radiomic signature with DE genes feature selection achieved better performance for DFS prediction than without. The nomogram incorporating the radiomic signature and lymph nodal status significantly stratified patients into high and low-risk groups for DFS (p < 0.001). The areas under the curve (AUCs) for predicting 5-year DFS were 0.912 in the training set, 0.852 in the internal test set, 0.769 in the external test set.

CONCLUSIONS : Genomics association was useful for radiomic feature selection. The established radiomic signature was prognostic for DFS. The radiomic nomogram could provide a valuable prediction for individualized long-term survival.

Xie Chen-Yi, Hu Yi-Huai, Ho Joshua Wing-Kei, Han Lu-Jun, Yang Hong, Wen Jing, Lam Ka-On, Wong Ian Yu-Hong, Law Simon Ying-Kit, Chiu Keith Wan-Hang, Fu Jian-Hua, Vardhanabhuti Varut

2021-Apr-29

esophageal squamous cell carcinoma, neoadjuvant chemoradiotherapy, prognosis, radiogenomic

General General

Detecting Topic and Sentiment Trends in Physician Rating Websites: Analysis of Online Reviews Using 3-Wave Datasets.

In International journal of environmental research and public health ; h5-index 73.0

(1) Background: Physician rating websites (PRWs) are a rich resource of information where individuals learn other people response to various health problems. The current study aims to investigate and analyze the people top concerns and sentiment dynamics expressed in physician online reviews (PORs). (2) Methods: Text data were collected from four U.S.-based PRWs during the three time periods of 2018, 2019 and 2020. Based on the dynamic topic modeling, hot topics related to different aspects of healthcare were identified. Following the hybrid approach of aspect-based sentiment analysis, the social network of prevailing topics was also analyzed whether people expressed positive, neutral or negative sentiments in PORs. (3) Results: The study identified 30 dominant topics across three different stages which lead toward four key findings. First, topics discussed in Stage III were quite different from the earlier two stages due to the COVID-19 outbreak. Second, based on the keyword co-occurrence analysis, the most prevalent keywords in all three stages were related to the treatment, questions asked by patients, communication problem, patients' feelings toward the hospital environment, disease symptoms, time spend with patients and different issues related to the COVID-19 (i.e., pneumonia, death, spread and cases). Third, topics related to the provider service quality, hospital servicescape and treatment cost were the most dominant topics in Stages I and II, while the quality of online information regarding COVID-19 and government countermeasures were the most dominant topics in Stage III. Fourth, when zooming into the topic-based sentiments analysis, hot topics in Stage I were mostly positive (joy be the dominant emotion), then negative (disgust be the dominant emotion) in Stage II. Furthermore, sentiments in the initial period of Stage III (COVID-19) were negative (anger be the dominant emotion), then transformed into positive (trust be the dominant emotion) later. The findings also revealed that the proposed method outperformed the conventional machine learning models in analyzing topic and sentiment dynamics expressed in PRWs. (4) Conclusions: Methodologically, this research demonstrates the ability and importance of computational techniques for analyzing large corpora of text and complementing conventional social science approaches.

Shah Adnan Muhammad, Naqvi Rizwan Ali, Jeong Ok-Ran

2021-Apr-29

COVID-19, discrete emotions, online reviews, sentiment analysis, text mining, topic modeling

Radiology Radiology

A Novel Method for COVID-19 Diagnosis Using Artificial Intelligence in Chest X-ray Images.

In Healthcare (Basel, Switzerland)

The Coronavirus disease 2019 (COVID-19) is an infectious disease spreading rapidly and uncontrollably throughout the world. The critical challenge is the rapid detection of Coronavirus infected people. The available techniques being utilized are body-temperature measurement, along with anterior nasal swab analysis. However, taking nasal swabs and lab testing are complex, intrusive, and require many resources. Furthermore, the lack of test kits to meet the exceeding cases is also a major limitation. The current challenge is to develop some technology to non-intrusively detect the suspected Coronavirus patients through Artificial Intelligence (AI) techniques such as deep learning (DL). Another challenge to conduct the research on this area is the difficulty of obtaining the dataset due to a limited number of patients giving their consent to participate in the research study. Looking at the efficacy of AI in healthcare systems, it is a great challenge for the researchers to develop an AI algorithm that can help health professionals and government officials automatically identify and isolate people with Coronavirus symptoms. Hence, this paper proposes a novel method CoVIRNet (COVID Inception-ResNet model), which utilizes the chest X-rays to diagnose the COVID-19 patients automatically. The proposed algorithm has different inception residual blocks that cater to information by using different depths feature maps at different scales, with the various layers. The features are concatenated at each proposed classification block, using the average-pooling layer, and concatenated features are passed to the fully connected layer. The efficient proposed deep-learning blocks used different regularization techniques to minimize the overfitting due to the small COVID-19 dataset. The multiscale features are extracted at different levels of the proposed deep-learning model and then embedded into various machine-learning models to validate the combination of deep-learning and machine-learning models. The proposed CoVIR-Net model achieved 95.7% accuracy, and the CoVIR-Net feature extractor with random-forest classifier produced 97.29% accuracy, which is the highest, as compared to existing state-of-the-art deep-learning methods. The proposed model would be an automatic solution for the assessment and classification of COVID-19. We predict that the proposed method will demonstrate an outstanding performance as compared to the state-of-the-art techniques being used currently.

Almalki Yassir Edrees, Qayyum Abdul, Irfan Muhammad, Haider Noman, Glowacz Adam, Alshehri Fahad Mohammed, Alduraibi Sharifa K, Alshamrani Khalaf, Alkhalik Basha Mohammad Abd, Alduraibi Alaa, Saeed M K, Rahman Saifur

2021-Apr-29

chest X-ray images, data analytics, feature extraction, healthcare, image processing, pandemic

General General

A Radiogenomics Ensemble to Predict EGFR and KRAS Mutations in NSCLC.

In Tomography (Ann Arbor, Mich.)

Lung cancer causes more deaths globally than any other type of cancer. To determine the best treatment, detecting EGFR and KRAS mutations is of interest. However, non-invasive ways to obtain this information are not available. Furthermore, many times there is a lack of big enough relevant public datasets, so the performance of single classifiers is not outstanding. In this paper, an ensemble approach is applied to increase the performance of EGFR and KRAS mutation prediction using a small dataset. A new voting scheme, Selective Class Average Voting (SCAV), is proposed and its performance is assessed both for machine learning models and CNNs. For the EGFR mutation, in the machine learning approach, there was an increase in the sensitivity from 0.66 to 0.75, and an increase in AUC from 0.68 to 0.70. With the deep learning approach, an AUC of 0.846 was obtained, and with SCAV, the accuracy of the model was increased from 0.80 to 0.857. For the KRAS mutation, both in the machine learning models (0.65 to 0.71 AUC) and the deep learning models (0.739 to 0.778 AUC), a significant increase in performance was found. The results obtained in this work show how to effectively learn from small image datasets to predict EGFR and KRAS mutations, and that using ensembles with SCAV increases the performance of machine learning classifiers and CNNs. The results provide confidence that as large datasets become available, tools to augment clinical capabilities can be fielded.

Moreno Silvia, Bonfante Mario, Zurek Eduardo, Cherezov Dmitry, Goldgof Dmitry, Hall Lawrence, Schabath Matthew

2021-Apr-29

CNN, EGFR, KRAS, NSCLC, ensembles, machine learning, radiogenomics

Radiology Radiology

Projecting COVID-19 disease severity in cancer patients using purposefully-designed machine learning.

In BMC infectious diseases ; h5-index 58.0

BACKGROUND : Accurately predicting outcomes for cancer patients with COVID-19 has been clinically challenging. Numerous clinical variables have been retrospectively associated with disease severity, but the predictive value of these variables, and how multiple variables interact to increase risk, remains unclear.

METHODS : We used machine learning algorithms to predict COVID-19 severity in 348 cancer patients at Memorial Sloan Kettering Cancer Center in New York City. Using only clinical variables collected on or before a patient's COVID-19 positive date (time zero), we sought to classify patients into one of three possible future outcomes: Severe-early (the patient required high levels of oxygen support within 3 days of being tested positive for COVID-19), Severe-late (the patient required high levels of oxygen after 3 days), and Non-severe (the patient never required oxygen support).

RESULTS : Our algorithm classified patients into these classes with an area under the receiver operating characteristic curve (AUROC) ranging from 70 to 85%, significantly outperforming prior methods and univariate analyses. Critically, classification accuracy is highest when using a potpourri of clinical variables - including basic patient information, pre-existing diagnoses, laboratory and radiological work, and underlying cancer type - suggesting that COVID-19 in cancer patients comes with numerous, combinatorial risk factors.

CONCLUSIONS : Overall, we provide a computational tool that can identify high-risk patients early in their disease progression, which could aid in clinical decision-making and selecting treatment options.

Navlakha Saket, Morjaria Sejal, Perez-Johnston Rocio, Zhang Allen, Taur Ying

2021-May-04

COVID-19, Cancer, Clinical machine learning, Infectious diseases, Predictive modeling

General General

The Coronavirus Network Explorer: mining a large-scale knowledge graph for effects of SARS-CoV-2 on host cell function.

In BMC bioinformatics

BACKGROUND : Leveraging previously identified viral interactions with human host proteins, we apply a machine learning-based approach to connect SARS-CoV-2 viral proteins to relevant host biological functions, diseases, and pathways in a large-scale knowledge graph derived from the biomedical literature. Our goal is to explore how SARS-CoV-2 could interfere with various host cell functions, and to identify drug targets amongst the host genes that could potentially be modulated against COVID-19 by repurposing existing drugs. The machine learning model employed here involves gene embeddings that leverage causal gene expression signatures curated from literature. In contrast to other network-based approaches for drug repurposing, our approach explicitly takes the direction of effects into account, distinguishing between activation and inhibition.

RESULTS : We have constructed 70 networks connecting SARS-CoV-2 viral proteins to various biological functions, diseases, and pathways reflecting viral biology, clinical observations, and co-morbidities in the context of COVID-19. Results are presented in the form of interactive network visualizations through a web interface, the Coronavirus Network Explorer (CNE), that allows exploration of underlying experimental evidence. We find that existing drugs targeting genes in those networks are strongly enriched in the set of drugs that are already in clinical trials against COVID-19.

CONCLUSIONS : The approach presented here can identify biologically plausible hypotheses for COVID-19 pathogenesis, explicitly connected to the immunological, virological and pathological observations seen in SARS-CoV-2 infected patients. The discovery of repurposable drugs is driven by prior knowledge of relevant functional endpoints that reflect known viral biology or clinical observations, therefore suggesting potential mechanisms of action. We believe that the CNE offers relevant insights that go beyond more conventional network approaches, and can be a valuable tool for drug repurposing. The CNE is available at https://digitalinsights.qiagen.com/coronavirus-network-explorer .

Krämer Andreas, Billaud Jean-Noël, Tugendreich Stuart, Shiffman Dan, Jones Martin, Green Jeff

2021-May-03

COVID-19, Drug repurposing, Knowledge graph, Network biology

General General

Using deep learning for acoustic event classification: The case of natural disasters.

In The Journal of the Acoustical Society of America

This study proposes a sound classification model for natural disasters. Deep learning techniques, a convolutional neural network (CNN) and long short-term memory (LSTM), were used to train two individual classifiers. The study was conducted using a dataset acquired online and truncated at 0.1 s to obtain a total of 12 937 sound segments. The result indicated that acoustic signals are effective for classifying natural disasters using machine learning techniques. The classifiers serve as an alternative effective approach to disaster classification. The CNN model obtained a classification accuracy of 99.96%, whereas the LSTM obtained an accuracy of 99.90%. The misclassification rates obtained in this study for the CNN and LSTM classifiers (i.e., 0.4% and 0.1%, respectively) suggest less classification errors when compared to existing studies. Future studies may investigate how to implement such classifiers for the early detection of natural disasters in real time.

Ekpezu Akon O, Wiafe Isaac, Katsriku Ferdinand, Yaokumah Winfred

2021-Apr

General General

Data augmentation for the classification of North Atlantic right whales upcalls.

In The Journal of the Acoustical Society of America

Passive acoustic monitoring (PAM) is a useful technique for monitoring marine mammals. However, the quantity of data collected through PAM systems makes automated algorithms for detecting and classifying sounds essential. Deep learning algorithms have shown great promise in recent years, but their performance is limited by the lack of sufficient amounts of annotated data for training the algorithms. This work investigates the benefit of augmenting training datasets with synthetically generated samples when training a deep neural network for the classification of North Atlantic right whale (Eubalaena glacialis) upcalls. We apply two recently proposed augmentation techniques, SpecAugment and Mixup, and show that they improve the performance of our model considerably. The precision is increased from 86% to 90%, while the recall is increased from 88% to 93%. Finally, we demonstrate that these two methods yield a significant improvement in performance in a scenario of data scarcity, where few training samples are available. This demonstrates that data augmentation can reduce the annotation effort required to achieve a desirable performance threshold.

Padovese Bruno, Frazao Fabio, Kirsebom Oliver S, Matwin Stan

2021-Apr

Radiology Radiology

Predicting intraventricular hemorrhage growth with a machine learning-based, radiomics-clinical model.

In Aging ; h5-index 49.0

We constructed a radiomics-clinical model to predict intraventricular hemorrhage (IVH) growth after spontaneous intracerebral hematoma. The model was developed using a training cohort (N=626) and validated with an independent testing cohort (N=270). Radiomics features and clinical predictors were selected using the least absolute shrinkage and selection operator (LASSO) method and multivariate analysis. The radiomics score (Rad-score) was calculated through linear combination of selected features multiplied by their respective LASSO coefficients. The support vector machine (SVM) method was used to construct the model. IVH growth was experienced by 13.4% and 13.7% of patients in the training and testing cohorts, respectively. The Rad-score was associated with severe IVH and poor outcome. Independent predictors of IVH growth included hypercholesterolemia (odds ratio [OR], 0.12 [95%CI, 0.02-0.90]; p=0.039), baseline Graeb score (OR, 1.26 [95%CI, 1.16-1.36]; p<0.001), time to initial CT (OR, 0.70 [95%CI, 0.58-0.86]; p<0.001), international normalized ratio (OR, 4.27 [95%CI, 1.40, 13.0]; p=0.011), and Rad-score (OR, 2.3 [95%CI, 1.6-3.3]; p<0.001). In the training cohort, the model achieved an AUC of 0.78, sensitivity of 0.83, and specificity of 0.66. In the testing cohort, AUC, sensitivity, and specificity were 0.71, 0.81, and 0.64, respectively. This radiomics-clinical model thus has the potential to predict IVH growth.

Zhu Dong-Qin, Chen Qian, Xiang Yi-Lan, Zhan Chen-Yi, Zhang Ming-Yue, Chen Chao, Zhuge Qi-Chuan, Chen Wei-Jian, Yang Xiao-Ming, Yang Yun-Jun

2021-May-04

cerebral intraventricular hemorrhage, decision support techniques, machine learning, multidetector computed tomography, precision medicine

General General

An interpretable machine learning method for supporting ecosystem management: Application to species distribution models of freshwater macroinvertebrates.

In Journal of environmental management

Species distribution models (SDMs), in which species occurrences are related to a suite of environmental variables, have been used as a decision-making tool in ecosystem management. Complex machine learning (ML) algorithms that lack interpretability may hinder the use of SDMs for ecological explanations, possibly limiting the role of SDMs as a decision-support tool. To meet the growing demand of explainable MLs, several interpretable ML methods have recently been proposed. Among these methods, SHaply Additive exPlanation (SHAP) has drawn attention for its robust theoretical justification and analytical gains. In this study, the utility of SHAP was demonstrated by the application of SDMs of four benthic macroinvertebrate species. In addition to species responses, the dataset contained 22 environmental variables monitored at 436 sites across five major rivers of South Korea. A range of ML algorithms was employed for model development. Each ML model was trained and optimized using 10-fold cross-validation. Model evaluation based on the test dataset indicated strong model performance, with an accuracy of ≥0.7 in all evaluation metrics for all MLs and species. However, only the random forest algorithm showed a behavior consistent with the known ecology of the investigated species. SHAP presents an integrated framework in which local interpretations that incorporate local interaction effects are combined to represent the global model structure. Consequently, this framework offered a novel opportunity to assess the importance of variables in predicting species occurrence, not only across sites, but also for individual sites. Furthermore, removing interaction effects from variable importance values (SHAP values) clearly revealed non-linear species responses to variations in environmental variables, indicating the existence of ecological thresholds. This study provides guidelines for the use of a new interpretable method supporting ecosystem management.

Cha YoonKyung, Shin Jihoon, Go ByeongGeon, Lee Dae-Seong, Kim YoungWoo, Kim TaeHo, Park Young-Seuk

2021-May-01

EPT taxa, Interpretable machine learning, Macroinvertebrate, SHAP, Species distribution model, Tree-based model

Pathology Pathology

Automated EEG pathology detection based on different convolutional neural network models: Deep learning approach.

In Computers in biology and medicine

The brain electrical activity, recorded and materialized as electroencephalogram (EEG) signals, is known to be very useful in the diagnosis of brain-related pathology. However, manual examination of these EEG signals has various limitations, including time-consuming inspections, the need for highly trained neurologists, and the subjectiveness of the evaluation. Thus, an automated EEG pathology detection system would be helpful to assist neurologists to enhance the treatment procedure by making a quicker diagnosis and reducing error due to the human element. This work proposes the application of a time-frequency spectrum to convert the EEG signals onto the image domain. The spectrum images are then applied to the Convolutional Neural Network (CNN) to learn robust features that can aid the automatic detection of pathology and normal EEG signals. Three popular CNN in the form of the DenseNet, Inception-ResNet v2, and SeizureNet were employed. The extracted deep-learned features from the spectrum images are then passed onto the support vector machine (SVM) classifier. The effectiveness of the proposed approach was assessed using the publicly available Temple University Hospital (TUH) abnormal EEG corpus dataset, which is demographically balanced. The proposed SeizureNet-SVM-based system achieved state-of-the-art performance: accuracy, sensitivity, and specificity of 96.65%, 90.48%, and 100%, respectively. The results show that the proposed framework may serve as a diagnostic tool to assist clinicians in the detection of EEG pathology for early treatment.

Bajpai Rishabh, Yuvaraj Rajamanickam, Prince A Amalin

2021-Apr-25

Abnormal EEG corpus, Convolutional neural network, Diagnosis, EEG, Pathology, Support vector machine

General General

Searching local order parameters to classify water structures of ice Ih, Ic, and liquid.

In The Journal of chemical physics

Identifying molecular structures of water and ice helps reveal the chemical nature of liquid and solid water. Real-space geometrical information on molecular systems can be precisely obtained from molecular simulations, but classifying the resulting structure is a non-trivial task. Order parameters are ordinarily introduced to effectively distinguish different structures. Many order parameters have been developed for various kinds of structures, such as body-centered cubic, face-centered cubic, hexagonal close-packed, and liquid. Order parameters for water have also been suggested but need further study. There has been no thorough investigation of the classification capability of many existing order parameters. In this work, we investigate the capability of 493 order parameters to classify the three structures of ice: Ih, Ic, and liquid. A total of 159 767 496 combinations of the order parameters are also considered. The investigation is automatically and systematically performed by machine learning. We find the best set of two bond-orientational order parameters, Q4 and Q8, to distinguish the three structures with high accuracy and robustness. A set of three order parameters is also suggested for better accuracy.

Doi Hideo, Takahashi Kazuaki Z, Aoyagi Takeshi

2021-Apr-28

General General

Correction: A Patient Journey Map to Improve the Home Isolation Experience of Persons With Mild COVID-19: Design Research for Service Touchpoints of Artificial Intelligence in eHealth.

In JMIR medical informatics ; h5-index 23.0

[This corrects the article DOI: 10.2196/23238.].

He Qian, Du Fei, Simonse Lianne W L

2021-May-04

Ophthalmology Ophthalmology

Machine learning in diagnosis and disability prediction of multiple sclerosis using optical coherence tomography.

In Computers in biology and medicine

BACKGROUND : Multiple sclerosis (MS) is a neurodegenerative disease that affects the central nervous system, especially the brain, spinal cord, and optic nerve. Diagnosis of this disease is a very complex process and generally requires a lot of time. In addition, treatments are applied without any information on the disability course in each MS patient. For these two reasons, the objective of this study was to improve the MS diagnosis and predict the long-term course of disability in MS patients based on clinical data and retinal nerve fiber layer (RNFL) thickness, measured by optical coherence tomography (OCT).

MATERIAL AND METHODS : A total of 104 healthy controls and 108 MS patients, 82 of whom had a 10-year follow-up, were enrolled. Classification algorithms such as multiple linear regression (MLR), support vector machines (SVM), decision tree (DT), k-nearest neighbours (k-NN), Naïve Bayes (NB), ensemble classifier (EC) and long short-term memory (LSTM) recurrent neural network were tested to develop two predictive models: MS diagnosis model and MS disability course prediction model.

RESULTS : For MS diagnosis, the best result was obtained using EC (accuracy: 87.7%; sensitivity: 87.0%; specificity: 88.5%; precision: 88.7%; AUC: 0.8775). In line with this good performance, the accuracy was 85.4% using k-NN and 84.4% using SVM. And, for long-term prediction of MS disability course, LSTM recurrent neural network was the most appropriate classifier (accuracy: 81.7%; sensitivity: 81.1%; specificity: 82.2%; precision: 78.9%; AUC: 0.8165). The use of MLR, SVM and k-NN also showed a good performance (AUC ≥ 0.8).

CONCLUSIONS : This study demonstrated that machine learning techniques, using clinical and OCT data, can help establish an early diagnosis and predict the course of MS. This advance could help clinicians select more specific treatments for each MS patient. Therefore, our findings underscore the potential of RNFL thickness as a reliable MS biomarker.

Montolío Alberto, Martín-Gallego Alejandro, Cegoñino José, Orduna Elvira, Vilades Elisa, Garcia-Martin Elena, Palomar Amaya Pérez Del

2021-Apr-26

Expanded disability status scale, Machine learning, Multiple sclerosis, Optical coherence tomography, Retinal nerve fiber layer

Radiology Radiology

Can magnetic resonance imaging radiomics of the pancreas predict postoperative pancreatic fistula?

In European journal of radiology ; h5-index 47.0

OBJECTIVES : To evaluate whether a magnetic resonance imaging (MRI) radiomics-based machine learning classifier can predict postoperative pancreatic fistula (POPF) after pancreaticoduodenectomy (PD) and to compare its performance to T1 signal intensity ratio (T1 SIratio).

METHODS : Sixty-two patients who underwent 3 T MRI before PD between 2008 and 2018 were retrospectively analyzed. POPF was graded and split into clinically relevant POPF (CR-POPF) vs. biochemical leak or no POPF. On T1- and T2-weighted images, 2 regions of interest were placed in the pancreatic corpus and cauda. 173 radiomics features were extracted using pyRadiomics. Additionally, the pancreas-to-muscle T1 SIratio was measured. The dataset was augmented and split into training (70 %) and test sets (30 %). A Boruta algorithm was used for feature reduction. For prediction of CR-POPF models were built using a gradient-boosted tree (GBT) and logistic regression from the radiomics features, T1 SIratio and a combination of the two. Diagnostic accuracy of the models was compared using areas under the receiver operating characteristics curve (AUCs).

RESULTS : Five most important radiomics features were identified for prediction of CR-POPF. A GBT using these features achieved an AUC of 0.82 (95 % Confidence Interval [CI]: 0.74 - 0.89) when applied on the original (non-augmented) dataset. Using T1 SIratio, a GBT model resulted in an AUC of 0.75 (CI: 0.63 - 0.84) and a logistic regression model delivered an AUC of 0.75 (CI: 0.63 - 0.84). A GBT model combining radiomics features and T1 SIratio resulted in an AUC of 0.90 (CI 0.84 - 0.95).

CONCLUSION : MRI-radiomics with routine sequences provides promising prediction of CR-POPF.

Skawran Stephan M, Kambakamba Patryk, Baessler Bettina, von Spiczak Jochen, Kupka Michael, Müller Philip C, Moeckli Beat, Linecker Michael, Petrowsky Henrik, Reiner Caecilia S

2021-Apr-24

Anastomotic leak, Artificial intelligence, Magnetic resonance imaging, Pancreas, Pancreatic fistula

Ophthalmology Ophthalmology

Automated detection of glaucoma with interpretable machine learning using clinical data and multi-modal retinal images.

In American journal of ophthalmology ; h5-index 67.0

PURPOSE : To develop a multi-modal model to automate glaucoma detection.

DESIGN : Development of a machine-learning glaucoma detection model.

METHODS : We selected a study cohort from the UK Biobank dataset with 1193 eyes of 863 healthy subjects and 1283 eyes of 771 subjects with glaucoma. We trained a multi-modal model that combines multiple deep neural nets, trained on macular optical coherence tomography volumes and color fundus photos, with demographic and clinical data. We performed an interpretability analysis to identify features the model relied on to detect glaucoma. We determined the importance of different features in detecting glaucoma using interpretable machine learning methods. We also evaluated the model on subjects who did not have a diagnosis of glaucoma on the day of imaging but were later diagnosed (progress-to-glaucoma, PTG).

RESULTS : Results show that a multi-modal model that combines imaging with demographic and clinical features is highly accurate (AUC 0.97). Interpretation of this model highlights biological features known to be related to the disease, such as age, intraocular pressure, and optic disc morphology. Our model also points to previously unknown or disputed features, such as pulmonary function and retinal outer layers. Accurate prediction in PTG highlights variables that change with progression to glaucoma - age and pulmonary function.

CONCLUSIONS : The accuracy of our model suggests distinct sources of information in each imaging modality and in the different clinical and demographic variables. Interpretable machine learning methods elucidate subject-level prediction and help uncover the factors that lead to accurate predictions, pointing to potential disease mechanisms or variables related to the disease.

Mehta Parmita, Petersen Christine A, Wen Joanne C, Banitt Michael R, Chen Philip P, Bojikian Karine D, Egan Catherine, Lee Su-In, Balazinska Magdalena, Lee Aaron Y, Rokem Ariel

2021-May-01

Public Health Public Health

Cause-specific mortality prediction in older residents of São Paulo, Brazil: a machine learning approach.

In Age and ageing ; h5-index 55.0

BACKGROUND : Populational ageing has been increasing in a remarkable rate in developing countries. In this scenario, preventive strategies could help to decrease the burden of higher demands for healthcare services. Machine learning algorithms have been increasingly applied for identifying priority candidates for preventive actions, presenting a better predictive performance than traditional parsimonious models.

METHODS : Data were collected from the Health, Well Being and Aging (SABE) Study, a representative sample of older residents of São Paulo, Brazil. Machine learning algorithms were applied to predict death by diseases of respiratory system (DRS), diseases of circulatory system (DCS), neoplasms and other specific causes within 5 years, using socioeconomic, demographic and health features. The algorithms were trained in a random sample of 70% of subjects, and then tested in the other 30% unseen data.

RESULTS : The outcome with highest predictive performance was death by DRS (AUC-ROC = 0.89), followed by the other specific causes (AUC-ROC = 0.87), DCS (AUC-ROC = 0.67) and neoplasms (AUC-ROC = 0.52). Among only the 25% of individuals with the highest predicted risk of mortality from DRS were included 100% of the actual cases. The machine learning algorithms with the highest predictive performance were light gradient boosted machine and extreme gradient boosting.

CONCLUSION : The algorithms had a high predictive performance for DRS, but lower for DCS and neoplasms. Mortality prediction with machine learning can improve clinical decisions especially regarding targeted preventive measures for older individuals.

do Nascimento Carla Ferreira, Dos Santos Hellen Geremias, de Moraes Batista André Filipe, Roman Lay Alejandra Andrea, Duarte Yeda Aparecida Oliveira, Chiavegatto Filho Alexandre Dias Porto

2021-May-03

machine learning, mortality, older people, prediction modelling

General General

Chronic stress in practice assistants: An analytic approach comparing four machine learning classifiers with a standard logistic regression model.

In PloS one ; h5-index 176.0

BACKGROUND : Occupational stress is associated with adverse outcomes for medical professionals and patients. In our cross-sectional study with 136 general practices, 26.4% of 550 practice assistants showed high chronic stress. As machine learning strategies offer the opportunity to improve understanding of chronic stress by exploiting complex interactions between variables, we used data from our previous study to derive the best analytic model for chronic stress: four common machine learning (ML) approaches are compared to a classical statistical procedure.

METHODS : We applied four machine learning classifiers (random forest, support vector machine, K-nearest neighbors', and artificial neural network) and logistic regression as standard approach to analyze factors contributing to chronic stress in practice assistants. Chronic stress had been measured by the standardized, self-administered TICS-SSCS questionnaire. The performance of these models was compared in terms of predictive accuracy based on the 'operating area under the curve' (AUC), sensitivity, and positive predictive value.

FINDINGS : Compared to the standard logistic regression model (AUC 0.636, 95% CI 0.490-0.674), all machine learning models improved prediction: random forest +20.8% (AUC 0.844, 95% CI 0.684-0.843), artificial neural network +12.4% (AUC 0.760, 95% CI 0.605-0.777), support vector machine +15.1% (AUC 0.787, 95% CI 0.634-0.802), and K-nearest neighbours +7.1% (AUC 0.707, 95% CI 0.556-0.735). As best prediction model, random forest showed a sensitivity of 99% and a positive predictive value of 79%. Using the variable frequencies at the decision nodes of the random forest model, the following five work characteristics influence chronic stress: too much work, high demand to concentrate, time pressure, complicated tasks, and insufficient support by practice leaders.

CONCLUSIONS : Regarding chronic stress prediction, machine learning classifiers, especially random forest, provided more accurate prediction compared to classical logistic regression. Interventions to reduce chronic stress in practice personnel should primarily address the identified workplace characteristics.

Bozorgmehr Arezoo, Thielmann Anika, Weltermann Birgitta

2021

General General

Analysis of facial expressions in response to basic taste stimuli using artificial intelligence to predict perceived hedonic ratings.

In PloS one ; h5-index 176.0

Taste stimuli can induce a variety of physiological reactions depending on the quality and/or hedonics (overall pleasure) of tastants, for which objective methods have long been desired. In this study, we used artificial intelligence (AI) technology to analyze facial expressions with the aim of assessing its utility as an objective method for the evaluation of food and beverage hedonics compared with conventional subjective (perceived) evaluation methods. The face of each participant (10 females; age range, 21-22 years) was photographed using a smartphone camera a few seconds after drinking 10 different solutions containing five basic tastes with different hedonic tones. Each image was then uploaded to an AI application to achieve outcomes for eight emotions (surprise, happiness, fear, neutral, disgust, sadness, anger, and embarrassment), with scores ranging from 0 to 100. For perceived evaluations, each participant also rated the hedonics of each solution from -10 (extremely unpleasant) to +10 (extremely pleasant). Based on these, we then conducted a multiple linear regression analysis to obtain a formula to predict perceived hedonic ratings. The applicability of the formula was examined by combining the emotion scores with another 11 taste solutions obtained from another 12 participants of both genders (age range, 22-59 years). The predicted hedonic ratings showed good correlation and concordance with the perceived ratings. To our knowledge, this is the first study to demonstrate a model that enables the prediction of hedonic ratings based on emotional facial expressions to food and beverage stimuli.

Yamamoto Takashi, Mizuta Haruno, Ueji Kayoko

2021

General General

Correction: A Patient Journey Map to Improve the Home Isolation Experience of Persons With Mild COVID-19: Design Research for Service Touchpoints of Artificial Intelligence in eHealth.

In JMIR medical informatics ; h5-index 23.0

[This corrects the article DOI: 10.2196/23238.].

He Qian, Du Fei, Simonse Lianne W L

2021-May-04

General General

Mutual-Prototype Adaptation for Cross-Domain Polyp Segmentation.

In IEEE journal of biomedical and health informatics

Accurate segmentation of the polyps from colonoscopy images provides useful information for the diagnosis and treatment of colorectal cancer. Despite deep learning methods advance automatic polyp segmentation, their performance often degrades when applied to new data acquired from different scanners or sequences (target domain). As manual annotation is tedious and labor-intensive for every new target domain, leveraging knowledge learned from the labeled source domain to promote the performance in the unlabeled target domain is highly demanded. In this work, we propose a mutual-prototype adaptation network to eliminate domain shifts in multi-centers and multi-devices colonoscopy images. We first devise a mutual-prototype alignment (MPA) module with the prototype relation function to refine features through self-domain and cross-domain information in a coarse-to-fine process. Then two auxiliary modules: progressive self-training (PST) and disentangled reconstruction (DR) are proposed to improve the segmentation performance. The PST module selects reliable pseudo labels through a novel uncertainty guided self-training loss to obtain accurate prototypes in the target domain. The DR module reconstructs original images jointly utilizing prediction results and private prototypes to maintain semantic consistency and provide complement supervision information. We extensively evaluate the proposed model in polyp segmentation performance on three conventional colonoscopy datasets: CVC-DB, Kvasir-SEG, and ETIS-Larib. The comprehensive experimental results demonstrate that the proposed model outperforms other state-of-the-art methods.

Yang Chen, Guo Xiaoqing, Zhu Meilu, Ibragimov Bulat, Yuan Yixuan

2021-May-04

General General

VolumeNet: A Lightweight Parallel Network for Super-Resolution of MR and CT Volumetric Data.

In IEEE transactions on image processing : a publication of the IEEE Signal Processing Society

Deep learning-based super-resolution (SR) techniques have generally achieved excellent performance in the computer vision field. Recently, it has been proven that three-dimensional (3D) SR for medical volumetric data delivers better visual results than conventional two-dimensional (2D) processing. However, deepening and widening 3D networks increases training difficulty significantly due to the large number of parameters and small number of training samples. Thus, we propose a 3D convolutional neural network (CNN) for SR of medical volumetric data called ParallelNet using parallel connections. We construct a parallel connection structure based on the group convolution and feature aggregation to build a 3D CNN that is as wide as possible with few parameters. As a result, the model thoroughly learns more feature maps with larger receptive fields. In addition, to further improve accuracy, we present an efficient version of ParallelNet (called VolumeNet), which reduces the number of parameters and deepens ParallelNet using a proposed lightweight building block module called the Queue module. Unlike most lightweight CNNs based on depthwise convolutions, the Queue module is primarily constructed using separable 2D cross-channel convolutions. As a result, the number of network parameters and computational complexity can be reduced significantly while maintaining accuracy due to full channel fusion. Experimental results demonstrate that the proposed VolumeNet significantly reduces the number of model parameters and achieves high precision results compared to state-of-the-art methods.

Li Yinhao, Iwamoto Yutaro, Lin Lanfen, Xu Rui, Chen Yen-Wei

2021-May-04

General General

Mutual Information Regularized Feature-level Frankenstein for Discriminative Recognition.

In IEEE transactions on pattern analysis and machine intelligence ; h5-index 127.0

Deep learning recognition approaches can potentially perform better if we can extract a discriminative representation that controllably separates nuisance factors. In this paper, we propose a novel approach to explicitly enforce the extracted discriminative representation d, extracted latent variation l (e,g., background, unlabeled nuisance attributes), and semantic variation label vector s (e.g., labeled expressions/pose) to be independent and complementary to each other. We can cast this problem as an adversarial game in the latent space of an auto-encoder. Specifically, with the to-be-disentangled s, we propose to equip an end-to-end conditional adversarial network with the ability to decompose an input sample into d and l. However, we argue that maximizing the cross-entropy loss of semantic variation prediction from d is not sufficient to remove the impact of s from d, and that the uniform-target and entropy regularization are necessary. A collaborative mutual information regularization framework is further proposed to avoid unstable adversarial training. It is able to minimize the differentiable mutual information between the variables to enforce independence. The proposed discriminative representation inherits the desired tolerance property guided by prior knowledge of the task. Our proposed framework achieves top performance on diverse recognition tasks.

Liu Xiaofeng, Chao Yang, You Jane J, Kuo C-C Jay, Vijayakumar Bhagavatula

2021-May-04

Public Health Public Health

Using machine learning to develop a novel COVID-19 Vulnerability Index (C19VI).

In The Science of the total environment

COVID-19 is now one of the most leading causes of death in the United States (US). Systemic health, social and economic disparities have put the minorities and economically poor communities at a higher risk than others. There is an immediate requirement to develop a reliable measure of county-level vulnerabilities that can capture the heterogeneity of vulnerable communities. This study reports a COVID-19 Vulnerability Index (C19VI) for identifying and mapping vulnerable counties. We proposed a Random Forest machine learning-based vulnerability model using CDC's sociodemographic and COVID-19-specific themes. An innovative 'COVID-19 Impact Assessment' algorithm was also developed for evaluating severity of the pandemic and to train the vulnerability model. Developed C19VI was statistically validated and compared with the CDC COVID-19 Community Vulnerability Index (CCVI). Finally, using C19VI and the census data, we explored racial inequalities and economic disparities in COVID-19 health outcomes. Our index indicates that 575 counties (45 million people) fall into the 'very high' vulnerability class, 765 counties (66 million people) in the 'high' vulnerability class, and 1435 counties (204 million people) in the 'moderate' or 'low' vulnerability class. Only 367 counties (20 million people) were found as 'very low' vulnerable areas. Furthermore, C19VI reveals that 524 counties with a racial minority population higher than 13% and 420 counties with poverty higher than 20% are in the 'very high' or 'high' vulnerability classes. The C19VI aims at helping public health officials and disaster management agencies to develop effective mitigation strategies especially for the disproportionately impacted communities.

Tiwari Anuj, Dadhania Arya V, Ragunathrao Vijay Avin Balaji, Oliveira Edson R A

2021-Jun-15

COVID-19, Disproportionate COVID-19, Machine learning, Racial minority, Vulnerability modeling

General General

News Sentiment Informed Time-series Analyzing AI (SITALA) to curb the spread of COVID-19 in Houston.

In Expert systems with applications

Coronavirus disease (COVID-19) has evolved into a pandemic with many unknowns. Houston, located in the Harris County of Texas, is becoming the next hotspot of this pandemic. With a severe decline in international and inter-state travel, a model at the county level is needed as opposed to the state or country level. Existing approaches have a few drawbacks. Firstly, the data used is the number of COVID-19 positive cases instead of positivity. The former is a function of the number of tests carried out while the number of tests normalizes the latter. Positivity gives a better picture of the spread of this pandemic as, with time, more tests are being administered. Positivity under 5% has been desired for the reopening of businesses to almost 100% capacity. Secondly, the data used by models like SEIRD (Susceptible, Exposed, Infectious, Recovered, and Deceased) lacks information about the sentiment of people concerning coronavirus. Thirdly, models that make use of social media posts might have too much noise and misinformation. On the other hand, news sentiment can capture long-term effects of hidden variables like public policy, opinions of local doctors, and disobedience of state-wide mandates. The present study introduces a new artificial intelligence (i.e., AI) model, viz., Sentiment Informed Time-series Analyzing AI (SITALA), trained on COVID-19 test positivity data and news sentiment from over 2750 news articles for Harris county. The news sentiment was obtained using IBM Watson Discovery News. SITALA is inspired by Google-Wavenet architecture and makes use of TensorFlow. The mean absolute error for the training dataset of 66 consecutive days is 2.76, and that for the test dataset of 22 consecutive days is 9.6. A cone of uncertainty is provided within which future COVID-19 test positivity has been shown to fall with high accuracy. The model predictions fare better than a published Bayesian-based SEIRD model. The model forecasts that in order to curb the spread of coronavirus in Houston, a sustained negative news sentiment (e.g., death count for COVID-19 will grow at an alarming rate in Houston if mask orders are not followed) will be desirable. Public policymakers may use SITALA to set the tone of the local policies and mandates.

Desai Prathamesh S

2021-Apr-29

Artificial Intelligence, COVID-19 Model, Deep Learning, News Sentiment, Pandemic Forecast, Public Policy

General General

Collective Reflective Equilibrium in Practice (CREP) and controversial novel technologies.

In Bioethics ; h5-index 25.0

In this paper, we investigate how data about public preferences may be used to inform policy around the use of controversial novel technologies, using public preferences about autonomous vehicles (AVs) as a case study. We first summarize the recent 'Moral Machine' study, which generated preference data from millions of people regarding how they think AVs should respond to emergency situations. We argue that while such preferences cannot be used to directly inform policy, they should not be disregarded. We defend an approach that we call 'Collective Reflective Equilibrium in Practice' (CREP). In CREP, data on public attitudes function as an input into a deliberative process that looks for coherence between attitudes, behaviours and competing ethical principles. We argue that in cases of reasonable moral disagreement, data on public attitudes should play a much greater role in shaping policies than in areas of ethical consensus. We apply CREP to some of the global preferences about AVs uncovered by the Moral Machines study. We intend this discussion both as a substantive contribution to the debate about the programming of ethical AVs, and as an illustration of how CREP works. We argue that CREP provides a principled way of using some public preferences as an input for policy, while justifiably disregarding others.

Savulescu Julian, Gyngell Christopher, Kahane Guy

2021-May-04

algorithm, artificial intelligence, bias, driverless cars, egalitarianism, ethical decision procedures, policy, reflective equilibrium, utilitarianism, veil of ignorance

General General

Contact-Free Pulse Signal Extraction from Human Face Videos: A Review and New Optimized Filtering Approach.

In Advances in experimental medicine and biology

In this chapter, we review methods for video-based heart monitoring, from classical signal processing approaches to modern deep learning methods. In addition, we propose a new method for learning an optimal filter that can overcome many of the problems that can affect classical approaches, such as light reflection and subject's movements, at a fraction of the training cost of deep learning approaches. Following the usual procedures for region of interest extraction and tracking, robust skin color estimation and signal pre-processing, we introduce a least-squares error optimal filter, learnt using an established training dataset to estimate the photoplethysmographic (PPG) signal more accurately from the measured color changes over time. This method not only improves the accuracy of heart rate measurement but also resulted in the extraction of a cleaner pulse signal, which could be integrated into many other useful applications such as human biometric recognition or recognition of emotional state. The method was tested on the DEAP dataset and showed improved performance over the best previous classical method on that dataset. The results obtained show that our proposed contact-free heart rate measurement method has significantly improved on existing methods.

Waqar Muhammad, Zwiggelaar Reyer, Tiddeman Bernard

2021

Deep learning, Optimal Fourier domain filter, Video photoplenthysmography (PPG)

General General

Harm Avoidance and Mobility During Middle Childhood and Adolescence among Hadza Foragers.

In Human nature (Hawthorne, N.Y.)

Cross-cultural sex differences in mobility and harm avoidance have been widely reported, often emphasizing fitness benefits of long-distance travel for males and high costs for females. Data emerging from adults in small-scale societies, however, are challenging the assumption that female mobility is restricted during reproduction. Such findings warrant further exploration of the ontogeny of mobility. Here, using a combination of machine-learning, mixed-effects linear regression, and GIS mapping, we analyze range size, daily distance traveled, and harm avoidance among Hadza foragers during middle childhood and adolescence. Distance traveled increased with age and, while male adolescents had the longest daily ranges, average daily distance traveled by each sex was similar. We found few age- or sex-related patterns in harm-avoidant responses and a high degree of individual variation. When queried on the same issues, children and their parents were often in alignment as to expectations pertaining to harm avoidance, and siblings tended to behave in similar ways. To the extent that sex differences in mobility did emerge, they were associated with ecological differences in physical threats associated with sex-specific foraging behaviors. Further, we found no strong association between harm avoidance and mobility. Young Hadza foragers of both sexes are highly mobile, regardless of how harm avoidant they are. Taken together, our findings indicate that the causal arrows between harm avoidance and mobility must be evaluated in ecologically specific frameworks where cultural expectations of juvenile mobility can be contextualized.

Crittenden Alyssa N, Farahani Alan, Herlosky Kristen N, Pollom Trevor R, Mabulla Ibrahim A, Ruginski Ian T, Cashdan Elizabeth

2021-May-04

Adolescence, Hadza, Harm avoidance, Middle childhood, Mobility

General General

Imagined character recognition through EEG signals using deep convolutional neural network.

In Medical & biological engineering & computing ; h5-index 32.0

Electroencephalography (EEG)-based brain computer interface (BCI) enables people to interact directly with computing devices through their brain signals. A BCI typically interprets EEG signals to reflect the user's intent or other mental activity. Motor imagery (MI) is a commonly used technique in BCIs where a user is asked to imagine moving certain part of the body such as a hand or a foot. By correctly interpreting the signal, one can perform a multitude of tasks such as controlling wheel chair, playing computer games, or even typing text. However, the use of motor-imagery-based BCIs outside the laboratory environment is limited due to the lack of their reliability. This work focuses on another kind of mental imagery, namely, the visual imagery (VI). VI is the manipulation of visual information that comes from memory. This work presents a deep convolutional neural network (DCNN)-based system for the recognition of visual/mental imagination of English alphabets so as to enable typing directly via brain signals. The DCNN learns to extract the spatial features hidden in the EEG signal. As opposed to many deep neural networks that use raw EEG signals for classification, this work transforms the raw signals into band powers using Morlet wavelet transformation. The proposed approach is evaluated on two publicly available benchmark MI-EEG datasets and a visual imagery dataset specifically collected for this work. The obtained results demonstrate that the proposed model performs better than the existing state-of-the-art methods for MI-EEG classification and yields an average accuracy of 99.45% on the two public MI-EEG datasets. The model also achieves an average recognition rate of 95.2% for the 26 English-language alphabets. Overall working of the proposed solution for imagined character recognition through EEG signals.

Ullah Sadiq, Halim Zahid

2021-May-04

Brain computer interfacing, Deep convolutional neural network, Deep learning, Supervised learning, Visual imagery

General General

The synergistic efficacy of hydroxychloroquine with methotrexate is accompanied by increased erythrocyte mean corpuscular volume.

In Rheumatology (Oxford, England)

OBJECTIVES : To determine whether concomitant hydroxychloroquine modulates the increase in erythrocyte mean corpuscular volume (MCV) caused by methotrexate therapy, and whether this is associated with improved clinical response in rheumatoid arthritis (RA).

METHODS : A retrospective observational analysis was conducted on two independent hospital datasets of biologic-naïve, early-RA patients who started oral methotrexate. Baseline characteristics, DAS28-ESR and monthly MCV after starting methotrexate were obtained. Conventional and machine-learning statistical approaches were applied to the discovery cohort (Cohort-1, 655 patients) and results validated using Cohort 2 (225 patients).

RESULTS : Hydroxychloroquine therapy with methotrexate was associated with a two-fold increase in the likelihood of response defined in this study as clinical remission or low disease activity at 6 months (p < 0.001). The improved clinical outcome of combination hydroxychloroquine and methotrexate therapy was associated with an accelerated rise in MCV from 2 months after commencing therapy. The increase in MCV at three months was equivalent to the contemporaneous reduction in the disease activity score (DAS28-ESR) in predicting clinical response at 6 months. Using latent class mixed modelling, five trajectories of MCV change over six months from baseline were identified. The odds ratio of response to treatment was 16.2 (95% CI 5.7 to 46.4, <ip > <0.001) in those receiving combination therapy classified within the MCV elevation >5fL class, which contained the most patients, compared to methotrexate alone.

CONCLUSION : Our data provides mechanistic insight into the synergistic clinical benefit of concomitant hydroxychloroquine with methotrexate, boosting the rise in MCV which could serve as a companion biomarker of treatment response.

Amin Shipa Muhammad Ruhu, Yeoh Su-Ann, Embleton-Thirsk Andrew, Mukerjee Dev, Ehrenstein Michael R

2021-May-03

biomarker, hydroxychloroquine, mean corpuscular change, methotrexate, synergistic

Public Health Public Health

Machine Learning and Natural Language Processing in Mental Health: Systematic Review.

In Journal of medical Internet research ; h5-index 88.0

BACKGROUND : Machine learning systems are part of the field of artificial intelligence that automatically learn models from data to make better decisions. Natural language processing (NLP), by using corpora and learning approaches, provides good performance in statistical tasks, such as text classification or sentiment mining.

OBJECTIVE : The primary aim of this systematic review was to summarize and characterize, in methodological and technical terms, studies that used machine learning and NLP techniques for mental health. The secondary aim was to consider the potential use of these methods in mental health clinical practice.

METHODS : This systematic review follows the PRISMA (Preferred Reporting Items for Systematic Review and Meta-analysis) guidelines and is registered with PROSPERO (Prospective Register of Systematic Reviews; number CRD42019107376). The search was conducted using 4 medical databases (PubMed, Scopus, ScienceDirect, and PsycINFO) with the following keywords: machine learning, data mining, psychiatry, mental health, and mental disorder. The exclusion criteria were as follows: languages other than English, anonymization process, case studies, conference papers, and reviews. No limitations on publication dates were imposed.

RESULTS : A total of 327 articles were identified, of which 269 (82.3%) were excluded and 58 (17.7%) were included in the review. The results were organized through a qualitative perspective. Although studies had heterogeneous topics and methods, some themes emerged. Population studies could be grouped into 3 categories: patients included in medical databases, patients who came to the emergency room, and social media users. The main objectives were to extract symptoms, classify severity of illness, compare therapy effectiveness, provide psychopathological clues, and challenge the current nosography. Medical records and social media were the 2 major data sources. With regard to the methods used, preprocessing used the standard methods of NLP and unique identifier extraction dedicated to medical texts. Efficient classifiers were preferred rather than transparent functioning classifiers. Python was the most frequently used platform.

CONCLUSIONS : Machine learning and NLP models have been highly topical issues in medicine in recent years and may be considered a new paradigm in medical research. However, these processes tend to confirm clinical hypotheses rather than developing entirely new information, and only one major category of the population (ie, social media users) is an imprecise cohort. Moreover, some language-specific features can improve the performance of NLP methods, and their extension to other languages should be more closely investigated. However, machine learning and NLP techniques provide useful information from unexplored data (ie, patients' daily habits that are usually inaccessible to care providers). Before considering It as an additional tool of mental health care, ethical issues remain and should be discussed in a timely manner. Machine learning and NLP methods may offer multiple perspectives in mental health research but should also be considered as tools to support clinical practice.

Le Glaz Aziliz, Haralambous Yannis, Kim-Dufor Deok-Hee, Lenca Philippe, Billot Romain, Ryan Taylor C, Marsh Jonathan, DeVylder Jordan, Walter Michel, Berrouiguet Sofian, Lemey Christophe

2021-May-04

artificial intelligence, data mining, machine learning, mental health, natural language processing, psychiatry

General General

Estimating Somatotype from a Single-camera 3D Body Scanning System.

In European journal of sport science

Somatotype is an approach to quantify body physique (shape and body composition). Somatotyping by manual measurement (the anthropometric method) or visual rating (the photoscopic method) needs technical expertise to minimize intra- and inter-observer errors. This study aims to develop machine learning models which enable automatic estimation of Heath-Carter somatotypes using a single-camera 3D scanning system. Single-camera 3D scanning was used to obtain 3D imaging data and computer vision techniques to extract features of body shape. Machine learning models were developed to predict participants' somatotypes from the extracted shape features. These predicted somatotypes were compared against manual measurement procedures. Data were collected from 46 participants and used as the training/validation set for model developing, whilst data collected from 17 participants were used as the test set for model evaluation. Evaluation tests showed that the 3D scanning methods enable accurate (mean error < 0.5; intraclass correlation coefficients >0.8) and precise (test-retest root mean square error < 0.5; intraclass correlation coefficients >0.8) somatotype predictions. This study shows that the 3D scanning methods could be used as an alternative to traditional somatotyping approaches after the current models improve with the large datasets.

Chiu Chuang-Yuan, Ciems Raimonds, Thelwell Michael, Bullas Alice, Choppin Simon

2021-May-04

3D analysis, Body composition, Measurement, Modeling, Technology

Radiology Radiology

Assessment of an AI Aid in Detection of Adult Appendicular Skeletal Fractures by Emergency Physicians and Radiologists: A Multicenter Cross-sectional Diagnostic Study.

In Radiology ; h5-index 91.0

Background The interpretation of radiographs suffers from an ever-increasing workload in emergency and radiology departments, while missed fractures represent up to 80% of diagnostic errors in the emergency department. Purpose To assess the performance of an artificial intelligence (AI) system designed to aid radiologists and emergency physicians in the detection and localization of appendicular skeletal fractures. Materials and Methods The AI system was previously trained on 60 170 radiographs obtained in patients with trauma. The radiographs were randomly split into 70% training, 10% validation, and 20% test sets. Between 2016 and 2018, 600 adult patients in whom multiview radiographs had been obtained after a recent trauma, with or without one or more fractures of shoulder, arm, hand, pelvis, leg, and foot, were retrospectively included from 17 French medical centers. Radiographs with quality precluding human interpretation or containing only obvious fractures were excluded. Six radiologists and six emergency physicians were asked to detect and localize fractures with (n = 300) and fractures without (n = 300) the aid of software highlighting boxes around AI-detected fractures. Aided and unaided sensitivity, specificity, and reading times were compared by means of paired Student t tests after averaging of performances of each reader. Results A total of 600 patients (mean age ± standard deviation, 57 years ± 22; 358 women) were included. The AI aid improved the sensitivity of physicians by 8.7% (95% CI: 3.1, 14.2; P = .003 for superiority) and the specificity by 4.1% (95% CI: 0.5, 7.7; P < .001 for noninferiority) and reduced the average number of false-positive fractures per patient by 41.9% (95% CI: 12.8, 61.3; P = .02) in patients without fractures and the mean reading time by 15.0% (95% CI: -30.4, 3.8; P = .12). Finally, stand-alone performance of a newer release of the AI system was greater than that of all unaided readers, including skeletal expert radiologists, with an area under the receiver operating characteristic curve of 0.94 (95% CI: 0.92, 0.96). Conclusion The artificial intelligence aid provided a gain of sensitivity (8.7% increase) and specificity (4.1% increase) without loss of reading speed. © RSNA, 2021 Online supplemental material is available for this article.

Duron Loïc, Ducarouge Alexis, Gillibert André, Lainé Julia, Allouche Christian, Cherel Nicolas, Zhang Zekun, Nitche Nicolas, Lacave Elise, Pourchot Aloïs, Felter Adrien, Lassalle Louis, Regnard Nor-Eddine, Feydy Antoine

2021-May-04

Radiology Radiology

AI-based Strategies to Reduce Workload in Breast Cancer Screening with Mammography and Tomosynthesis: A Retrospective Evaluation.

In Radiology ; h5-index 91.0

Background The workflow of breast cancer screening programs could be improved given the high workload and the high number of false-positive and false-negative assessments. Purpose To evaluate if using an artificial intelligence (AI) system could reduce workload without reducing cancer detection in breast cancer screening with digital mammography (DM) or digital breast tomosynthesis (DBT). Materials and Methods Consecutive screening-paired and independently read DM and DBT images acquired from January 2015 to December 2016 were retrospectively collected from the Córdoba Tomosynthesis Screening Trial. The original reading settings were single or double reading of DM or DBT images. An AI system computed a cancer risk score for DM and DBT examinations independently. Each original setting was compared with a simulated autonomous AI triaging strategy (the least suspicious examinations for AI are not human-read; the rest are read in the same setting as the original, and examinations not recalled by radiologists but graded as very suspicious by AI are recalled) in terms of workload, sensitivity, and recall rate. The McNemar test with Bonferroni correction was used for statistical analysis. Results A total of 15 987 DM and DBT examinations (which included 98 screening-detected and 15 interval cancers) from 15 986 women (mean age ± standard deviation, 58 years ± 6) were evaluated. In comparison with double reading of DBT images (568 hours needed, 92 of 113 cancers detected, 706 recalls in 15 987 examinations), AI with DBT would result in 72.5% less workload (P < .001, 156 hours needed), noninferior sensitivity (95 of 113 cancers detected, P = .38), and 16.7% lower recall rate (P < .001, 588 recalls in 15 987 examinations). Similar results were obtained for AI with DM. In comparison with the original double reading of DM images (222 hours needed, 76 of 113 cancers detected, 807 recalls in 15 987 examinations), AI with DBT would result in 29.7% less workload (P < .001), 25.0% higher sensitivity (P < .001), and 27.1% lower recall rate (P < .001). Conclusion Digital mammography and digital breast tomosynthesis screening strategies based on artificial intelligence systems could reduce workload up to 70%. Published under a CC BY 4.0 license.

Raya-Povedano José Luis, Romero-Martín Sara, Elías-Cabot Esperanza, Gubern-Mérida Albert, Rodríguez-Ruiz Alejandro, Álvarez-Benito Marina

2021-May-04

General General

Analyzing Dynamical Disorder for Charge Transport in Organic Semiconductors via Machine Learning.

In Journal of chemical theory and computation

Organic semiconductors are indispensable for today's display technologies in the form of organic light-emitting diodes (OLEDs) and further optoelectronic applications. However, organic materials do not reach the same charge carrier mobility as inorganic semiconductors, limiting the efficiency of devices. To find or even design new organic semiconductors with higher charge carrier mobility, computational approaches, in particular multiscale models, are becoming increasingly important. However, such models are computationally very costly, especially when large systems and long timescales are required, which is the case to compute static and dynamic energy disorder, i.e., the dominant factor to determine charge transport. Here, we overcome this drawback by integrating machine learning models into multiscale simulations. This allows us to obtain unprecedented insight into relevant microscopic materials properties, in particular static and dynamic disorder contributions for a series of application-relevant molecules. We find that static disorder and thus the distribution of shallow traps are highly asymmetrical for many materials, impacting widely considered Gaussian disorder models. We furthermore analyze characteristic energy level fluctuation times and compare them to typical hopping rates to evaluate the importance of dynamic disorder for charge transport. We hope that our findings will significantly improve the accuracy of computational methods used to predict application-relevant materials properties of organic semiconductors and thus make these methods applicable for virtual materials design.

Reiser Patrick, Konrad Manuel, Fediai Artem, Léon Salvador, Wenzel Wolfgang, Friederich Pascal

2021-May-04

Public Health Public Health

Development of a Random Forest model for forecasting allergenic pollen in North America.

In The Science of the total environment

Pollen allergies have negative impacts on health. Information about airborne pollen concentration can improve symptom management by guiding choices affecting timing of medicines and pollen exposure. Observations provide accurate pollen concentrations at point locations. However, in the contiguous United States and southern Canada (CUSSC), observations are sparse, and sampling is often seasonal, intermittent or both. Modeling pollen concentration can fill in the gaps with estimates where direct observations are unavailable and also provide much-needed forecasts. The goal of this study is to develop and evaluate statistical models that predict daily pollen concentrations using a machine learning Random Forest algorithm. To evaluate our methods, we made retrospective forecasts of four pollen types (Quercus, Cupressaceae, Ambrosia and Poaceae), each in one of four CUSSC locations. Meteorological and vegetation conditions were input to the models at city and regional scales. A data augmentation technique was investigated and found to improve model skill. Models were also developed to forecast pollen in locations where there are no observations. Forecast skill in these models were found to be greater than in previous models. Nevertheless, the skill is limited by the spatiotemporal resolution of the pollen observations.

Lo Fiona, Bitz Cecilia M, Hess Jeremy J

2021-Jun-15

Allergy, Hay fever, Machine learning, Prediction, Weather

Radiology Radiology

Projecting COVID-19 disease severity in cancer patients using purposefully-designed machine learning.

In BMC infectious diseases ; h5-index 58.0

BACKGROUND : Accurately predicting outcomes for cancer patients with COVID-19 has been clinically challenging. Numerous clinical variables have been retrospectively associated with disease severity, but the predictive value of these variables, and how multiple variables interact to increase risk, remains unclear.

METHODS : We used machine learning algorithms to predict COVID-19 severity in 348 cancer patients at Memorial Sloan Kettering Cancer Center in New York City. Using only clinical variables collected on or before a patient's COVID-19 positive date (time zero), we sought to classify patients into one of three possible future outcomes: Severe-early (the patient required high levels of oxygen support within 3 days of being tested positive for COVID-19), Severe-late (the patient required high levels of oxygen after 3 days), and Non-severe (the patient never required oxygen support).

RESULTS : Our algorithm classified patients into these classes with an area under the receiver operating characteristic curve (AUROC) ranging from 70 to 85%, significantly outperforming prior methods and univariate analyses. Critically, classification accuracy is highest when using a potpourri of clinical variables - including basic patient information, pre-existing diagnoses, laboratory and radiological work, and underlying cancer type - suggesting that COVID-19 in cancer patients comes with numerous, combinatorial risk factors.

CONCLUSIONS : Overall, we provide a computational tool that can identify high-risk patients early in their disease progression, which could aid in clinical decision-making and selecting treatment options.

Navlakha Saket, Morjaria Sejal, Perez-Johnston Rocio, Zhang Allen, Taur Ying

2021-May-04

COVID-19, Cancer, Clinical machine learning, Infectious diseases, Predictive modeling

General General

Attack-agnostic Adversarial Detection on Medical Data Using Explainable Machine Learning

ArXiv Preprint

Explainable machine learning has become increasingly prevalent, especially in healthcare where explainable models are vital for ethical and trusted automated decision making. Work on the susceptibility of deep learning models to adversarial attacks has shown the ease of designing samples to mislead a model into making incorrect predictions. In this work, we propose a model agnostic explainability-based method for the accurate detection of adversarial samples on two datasets with different complexity and properties: Electronic Health Record (EHR) and chest X-ray (CXR) data. On the MIMIC-III and Henan-Renmin EHR datasets, we report a detection accuracy of 77% against the Longitudinal Adversarial Attack. On the MIMIC-CXR dataset, we achieve an accuracy of 88%; significantly improving on the state of the art of adversarial detection in both datasets by over 10% in all settings. We propose an anomaly detection based method using explainability techniques to detect adversarial samples which is able to generalise to different attack methods without a need for retraining.

Matthew Watson, Noura Al Moubayed

2021-05-05

Surgery Surgery

Machine learning analysis: general features, requirements and cardiovascular applications.

In Minerva cardiology and angiology

Artificial intelligence represents the science which will probably change the future of medicine by solving actually challenging issues. In this special article, the general features of machine learning are discussed. First, a background explanation regarding the division of artificial intelligence, machine learning and deep learning is given and a focus on the structure of machine learning subgroups is shown. The traditional process of a machine learning analysis is described, starting from the collection of data, across features engineering, modelling and till the validation and deployment phase. Due to the several applications of machine learning performed in literature in the last decades and the lack of some guidelines, the need of a standardization for reporting machine learning analysis results emerged. Some possible standards for reporting machine learning results are identified and discussed deeply; these are related to study population (number of subjects), repeatability of the analysis, validation, results, comparison with current practice. The way to the use of machine learning in clinical practice is open and the hope is that, with emerging technology and advanced digital and computational tools, available from hospitalization and subsequently after discharge, it will also be possible, with the help of increasingly powerful hardware, to build assistance strategies useful in clinical practice.

Ricciardi Carlo, Cuocolo Renato, Megna Rosario, Cesarelli Mario, Petretta Mario

2021-May-04

Surgery Surgery

Intraoperative and postoperative complications in colorectal procedures: the role of continuous updating in medicine.

In Minerva surgery

Accepting surgical complications, especially those related to the learning curve, as unavoidable events in colorectal procedures, is like accepting to fly onboard an aircraft with a 10 to 20% chance of not arriving at final destination. Under this condition, it is very likely that the aviation industry and the concurrent reshaping of the world and of our lives would have not been possible in the absence of high reliability and reproducibility of safe flights. It's hard to imagine surgery without any intraoperative and/or postoperative complications. Nevertheless, there is a plenty of room for improvement by simply adopting what has been explicitly and scientifically demonstrated; training outside of the OR, usage of modern information technologies and application of evidence-based perioperative care protocols. Additionally, the possibility to objectively measure and monitor the technical and even non-technical skills and competencies of individual surgeons and even of OR teams through the application of structured and validated assessment tools can finally put an end to the self-referential, purely hierarchical, and indeed extremely unreliable process of being authorized or not to perform operations on patients. Last but not least, a wide range of new technologies spanning from augmented imaging modalities, virtual reality for intraoperative guidance, improved robotic manipulators, artificial intelligence to assist in preoperative patient specific risk assessment, and intraoperative decision-making has the potential to tackle several hidden roots of surgical complications.

Forgione Antonello, Guraya Salman Y, Diana Michele, Marescaux Jacques

2021-May-04

Surgery Surgery

Clinical validation of an artificial intelligence-enabled wound imaging mobile application in diabetic foot ulcers.

In International wound journal

There is a lifetime risk of 15% to 25% of development of diabetic foot ulcers (DFUs) in patients with diabetes mellitus. DFUs need to be followed up on and assessed for development of complications and/or resolution, which was traditionally performed using manual measurement. Our study aims to compare the intra- and inter-rater reliability of an artificial intelligence-enabled wound imaging mobile application (CARES4WOUNDS [C4W] system, Tetsuyu, Singapore) with traditional measurement. This is a prospective cross-sectional study on 28 patients with DFUs from June 2020 to January 2021. The main wound parameters assessed were length and width. For traditional manual measurement, area was calculated by overlaying traced wound on graphical paper. Intra- and inter-rater reliability was analysed using intra-class correlation statistics. A value of <0.5, 0.5-0.75, 0.75-0.9, and >0.9 indicates poor, moderate, good, and excellent reliability, respectively. Seventy-five wound episodes from 28 patients were collected and a total of 547 wound images were analysed in this study. The median wound area during the first clinic consultation and all wound episodes was 3.75 cm2 (interquartile range [IQR] 1.40-16.50) and 3.10 cm2 (IQR 0.60-14.84), respectively. There is excellent intra-rater reliability of C4W on three different image captures of the same wound (intra-rater reliability ranging 0.933-0.994). There is also excellent inter-rater reliability between three C4W devices for length (0.947), width (0.923), and area (0.965). Good inter-rater reliability for length, width, and area (range 0.825-0.934) was obtained between wound nurse measurement and each of the C4W devices. In conclusion, we obtained good inter-rater and intra-rater reliability of C4W measurements against traditional wound measurement. The C4W is a useful adjunct in monitoring DFU wound progress.

Chan Kai Siang, Chan Yam Meng, Tan Audrey Hui Min, Liang Shanying, Cho Yuan Teng, Hong Qiantai, Yong Enming, Chong Lester Rhan Chaen, Zhang Li, Tan Glenn Wei Leong, Chandrasekar Sadhana, Lo Zhiwen Joseph

2021-May-04

artificial intelligence, diabetic foot, foot ulcer, mobile applications, wound healing

Public Health Public Health

Using causal forests to assess heterogeneity in cost-effectiveness analysis.

In Health economics

We develop a method for data-driven estimation and analysis of heterogeneity in cost-effectiveness analyses (CEA) with experimental or observational individual-level data. Our implementation uses causal forests and cross-fitted augmented inverse probability weighted learning to estimate heterogeneity in incremental outcomes, costs and net monetary benefits, as well as other parameters relevant to CEA. We also show how the results can be visualized in relevant ways for the analysis of heterogeneity in CEA, such as using individual-level cost effectiveness planes. Using a simulated dataset and an R package implementing our methods, we show how the approach can be used to estimate the average cost-effectiveness in the entire sample or in subpopulations, explore and analyze the heterogeneity in incremental outcomes, costs and net monetary benefits (and their determinants), and learn policy rules from the data.

Bonander Carl, Svensson Mikael

2021-May-04

causal forest, cost-effectiveness analysis, machine learning, stratified analysis, treatment heterogeneity

Surgery Surgery

Molecular and morphological findings in a sample of oral surgery patients: What can we learn for multivariate concepts for age estimation?

In Journal of forensic sciences

It has already been proposed that a combined use of different molecular and morphological markers of aging in multivariate models may result in a greater accuracy of age estimation. However, such an approach can be complex and expensive, and not every combination may be useful. The significance and usefulness of combined analyses of D-aspartic acid in dentine, pentosidine in dentine, DNA methylation in buccal swabs at five genomic regions (PDE4C, RPA2, ELOVL2, DDO, and EDARADD), and third molar mineralization were tested by investigating a sample of 90 oral surgery patients. Machine learning models for age estimation were trained and evaluated, and the contribution of each parameter to multivariate models was tested by assessment of the predictor importance. For models based on D-aspartic acid, pentosidine, and the combination of both, mean absolute errors (MAEs) of 2.93, 3.41, and 2.68 years were calculated, respectively. The additional inclusion of the five DNAm markers did not improve the results. The sole DNAm-based model revealed a MAE of 4.14 years. In individuals under 28 years of age, the combination of the DNAm markers with the third molar mineralization stages reduced the MAE from 3.85 to 2.81 years. Our findings confirm that the combination of parameters in multivariate models may be very useful for age estimation. However, the inclusion of many parameters does not necessarily lead to better results. It is a task for future research to identify the best selection of parameters for the different requirements in forensic practice.

Siahaan Tatjana, Reckert Alexandra, Becker Julia, Eickhoff Simon B, Koop Barbara, Gündüz Tanju, Böhme Petra, Mayer Felix, Küppers Lisa, Wagner Wolfgang, Ritz-Timme Stefanie

2021-May-04

D-aspartic acid, DNA methylation, age estimation, multivariate models, pentosidine, tooth mineralization stages

Cardiology Cardiology

Assessment of causal association between thyroid function and lipid metabolism: a Mendelian randomization study.

In Chinese medical journal

BACKGROUND : Thyroid dysfunction is associated with cardiovascular diseases. However, the role of thyroid function in lipid metabolism remains partly unknown. The present study aimed to investigate the causal association between thyroid function and serum lipid metabolism via a genetic analysis termed Mendelian randomization (MR).

METHODS : The MR approach uses a genetic variant as the instrumental variable in epidemiological studies to mimic a randomized controlled trial. A two-sample MR was performed to assess the causal association, using summary statistics from the Atrial Fibrillation Genetics Consortium (n = 537,409) and the Global Lipids Genetics Consortium (n = 188,577). The clinical measures of thyroid function include thyrotropin (TSH), free triiodothyronine (FT3) and free thyroxine (FT4) levels, FT3:FT4 ratio and concentration of thyroid peroxidase antibodies (TPOAb). The serum lipid metabolism traits include total cholesterol (TC) and triglycerides, high-density lipoprotein, and low-density lipoprotein (LDL) levels. The MR estimate and MR inverse variance-weighted method were used to assess the association between thyroid function and serum lipid metabolism.

RESULTS : The results demonstrated that increased TSH levels were significantly associated with higher TC (β = 0.052, P = 0.002) and LDL (β = 0.041, P = 0.018) levels. In addition, the FT3:FT4 ratio was significantly associated with TC (β = 0.240, P = 0.033) and LDL (β = 0.025, P = 0.027) levels. However, no significant differences were observed between genetically predicted FT4 and TPOAb and serum lipids.

CONCLUSION : Taken together, the results of the present study suggest an association between thyroid function and serum lipid metabolism, highlighting the importance of the pituitary-thyroid-cardiac axis in dyslipidemia susceptibility.

Wang Jing-Jia, Zhuang Zhen-Huang, Shao Chun-Li, Yu Can-Qing, Wang Wen-Yao, Zhang Kuo, Meng Xiang-Bin, Gao Jun, Tian Jian, Zheng Ji-Lin, Huang Tao, Tang Yi-Da

2021-Apr-13

General General

Sniff-synchronized, gradient-guided olfactory search by freely-moving mice.

In eLife

For many organisms, searching for relevant targets such as food or mates entails active, strategic sampling of the environment. Finding odorous targets may be the most ancient search problem that motile organisms evolved to solve. While chemosensory navigation has been well characterized in micro-organisms and invertebrates, spatial olfaction in vertebrates is poorly understood. We have established an olfactory search assay in which freely-moving mice navigate noisy concentration gradients of airborne odor. Mice solve this task using concentration gradient cues and do not require stereo olfaction for performance. During task performance, respiration and nose movement are synchronized with tens of milliseconds precision. This synchrony is present during trials and largely absent during inter-trial intervals, suggesting that sniff-synchronized nose movement is a strategic behavioral state rather than simply a constant accompaniment to fast breathing. To reveal the spatiotemporal structure of these active sensing movements, we used machine learning methods to parse motion trajectories into elementary movement motifs. Motifs fall into two clusters, which correspond to investigation and approach states. Investigation motifs lock precisely to sniffing, such that the individual motifs preferentially occur at specific phases of the sniff cycle. The allocentric structure of investigation and approach indicate an advantage to sampling both sides of the sharpest part of the odor gradient, consistent with a serial sniff strategy for gradient sensing. This work clarifies sensorimotor strategies for mouse olfactory search and guides ongoing work into the underlying neural mechanisms.

Findley Teresa M, Wyrick David G, Cramer Jennifer L, Brown Morgan A, Holcomb Blake, Attey Robin, Yeh Dorian, Monasevitch Eric, Nouboussi Nelly, Cullen Isabelle, Songco Jeremea O, King Jared F, Ahmadian Yashar, Smear Matthew C

2021-May-04

mouse, neuroscience

General General

Machine Learning Approaches to Identify Factors Associated with Women's Vasomotor Symptoms Using General Hospital Data.

In Journal of Korean medical science

BACKGROUND : To analyze the factors associated with women's vasomotor symptoms (VMS) using machine learning.

METHODS : Data on 3,298 women, aged 40-80 years, who attended their general health check-up from January 2010 to December 2012 were obtained from Korea University Anam Hospital in Seoul, Korea. Five machine learning methods were applied and compared for the prediction of VMS, measured by the Menopause Rating Scale. Variable importance, the effect of a variable on model performance, was used for identifying the major factors associated with VMS.

RESULTS : In terms of the mean squared error, the random forest (0.9326) was much better than linear regression (12.4856) and artificial neural networks with one, two, and three hidden layers (1.5576, 1.5184, and 1.5833, respectively). Based on the variable importance from the random forest, the most important factors associated with VMS were age, menopause age, thyroid-stimulating hormone, and monocyte, triglyceride, gamma glutamyl transferase, blood urea nitrogen, cancer antigen 19-9, C-reactive protein, and low-density lipoprotein cholesterol levels. Indeed, the following variables were ranked within the top 20 in terms of variable importance: cancer antigen 125, total cholesterol, insulin, free thyroxine, forced vital capacity, alanine aminotransferase, forced expired volume in 1 second, height, homeostatic model assessment for insulin resistance, and carcinoembryonic antigen.

CONCLUSION : Machine learning provides an invaluable decision support system for the prediction of VMS. For managing VMS, comprehensive consideration is needed regarding thyroid function, lipid profile, liver function, inflammation markers, insulin resistance, monocyte count, cancer antigens, and lung function.

Ryu Ki Jin, Yi Kyong Wook, Kim Yong Jin, Shin Jung Ho, Hur Jun Young, Kim Tak, Seo Jong Bae, Lee Kwang Sig, Park Hyuntae

2021-May-03

Cancer Antigen, Hot Flashes, Menopause Age, Monocyte, Thyroid Stimulating Hormone, Vasomotor Symptoms

General General

Computational flow cytometry as a diagnostic tool in suspected-myelodysplastic syndromes.

In Cytometry. Part A : the journal of the International Society for Analytical Cytology

The diagnostic work-up of patients suspected for myelodysplastic syndromes is challenging and mainly relies on bone marrow morphology and cytogenetics. In this study, we developed and prospectively validated a fully computational tool for flow cytometry diagnostics in suspected-MDS. The computational diagnostic workflow consists of methods for pre-processing flow cytometry data, followed by a cell population detection method (FlowSOM) and a machine learning classifier (Random Forest). Based on a six tubes FC panel, the workflow obtained a 90% sensitivity and 93% specificity in an independent validation cohort. For practical advantages (e.g. reduced processing time and costs), a second computational diagnostic workflow was trained, solely based on the best performing single tube of the training cohort. This workflow obtained 97% sensitivity and 95% specificity in the prospective validation cohort. Both workflows outperformed the conventional, expert analyzed flow cytometry scores for diagnosis with respect to accuracy, objectivity and time investment (less than three minutes per patient). This article is protected by copyright. All rights reserved.

Duetz Carolien, Van Gassen Sofie, Westers Theresia M, van Spronsen Margot F, Bachas Costa, Saeys Yvan, van de Loosdrecht Arjan A

2021-May-03

Diagnostic test, Flow Cytometry, Hematological malignancies, Machine learning, Myelodysplastic syndromes (MDS)

Radiology Radiology

Diagnostic power of resting-state fMRI for detection of network connectivity in Alzheimer's disease and mild cognitive impairment: A systematic review.

In Human brain mapping

Resting-state fMRI (rs-fMRI) detects functional connectivity (FC) abnormalities that occur in the brains of patients with Alzheimer's disease (AD) and mild cognitive impairment (MCI). FC of the default mode network (DMN) is commonly impaired in AD and MCI. We conducted a systematic review aimed at determining the diagnostic power of rs-fMRI to identify FC abnormalities in the DMN of patients with AD or MCI compared with healthy controls (HCs) using machine learning (ML) methods. Multimodal support vector machine (SVM) algorithm was the commonest form of ML method utilized. Multiple kernel approach can be utilized to aid in the classification by incorporating various discriminating features, such as FC graphs based on "nodes" and "edges" together with structural MRI-based regional cortical thickness and gray matter volume. Other multimodal features include neuropsychiatric testing scores, DTI features, and regional cerebral blood flow. Among AD patients, the posterior cingulate cortex (PCC)/Precuneus was noted to be a highly affected hub of the DMN that demonstrated overall reduced FC. Whereas reduced DMN FC between the PCC and anterior cingulate cortex (ACC) was observed in MCI patients. Evidence indicates that the nodes of the DMN can offer moderate to high diagnostic power to distinguish AD and MCI patients. Nevertheless, various concerns over the homogeneity of data based on patient selection, scanner effects, and the variable usage of classifiers and algorithms pose a challenge for ML-based image interpretation of rs-fMRI datasets to become a mainstream option for diagnosing AD and predicting the conversion of HC/MCI to AD.

Ibrahim Buhari, Suppiah Subapriya, Ibrahim Normala, Mohamad Mazlyfarina, Hassan Hasyma Abu, Nasser Nisha Syed, Saripan M Iqbal

2021-May-04

“Alzheimers disease”, accuracy, classifiers, default mode network, functional MRI, machine learning

General General

Deep learning-based downscaling of tropospheric nitrogen dioxide using ground-level and satellite observations.

In The Science of the total environment

Air quality is one of the major issues within an urban area that affect people's living environment and health conditions. Existing observations are not adequate to provide a spatiotemporally comprehensive air quality information for vulnerable populations to plan ahead. Launched in 2017, TROPOspheric Monitoring Instrument (TROPOMI) provides a high spatial resolution (~5 km) tropospheric air quality measurement that captures the spatial variability of air pollution, but still limited by its daily overpass in the temporal dimension and relatively short historical records. Integrating with the hourly available AirNOW observations by ground-level discrete stations, we proposed and compared two deep learning methods that learn the relationship between the ground-level nitrogen dioxide (NO2) observation from AirNOW and the tropospheric NO2 column density from TROPOMI to downscale the daily NO2 to an hourly resolution. The input predictors include the locations of AirNOW stations, AirNOW NO2 observations, boundary layer height, other meteorological status, elevation, major roads, and power plants. The learned relationship can be used to produce NO2 emission estimates at the sub-urban scale on an hourly basis. The two methods include 1) an integrated method between inverse weighted distance and a feed forward neural network (IDW + DNN), and 2) a deep matrix network (DMN) that maps the discrete AirNOW observations directly to the distribution of TROPOMI observations. We further compared the accuracies of both models using different configurations of input predictors and validated their average Root Mean Squared Error (RMSE), average Mean Absolute Error (MAE) and the spatial distribution of errors. Results show that DMN generates more reliable NO2 estimates and captures a better spatial distribution of NO2 concentrations than the IDW + DNN model.

Yu Manzhu, Liu Qian

2021-Jun-15

AirNOW, Deep learning, Nitrogen dioxide, Spatial downscaling, Spatial interpolation, TROPOMI

General General

The Coronavirus Network Explorer: mining a large-scale knowledge graph for effects of SARS-CoV-2 on host cell function.

In BMC bioinformatics

BACKGROUND : Leveraging previously identified viral interactions with human host proteins, we apply a machine learning-based approach to connect SARS-CoV-2 viral proteins to relevant host biological functions, diseases, and pathways in a large-scale knowledge graph derived from the biomedical literature. Our goal is to explore how SARS-CoV-2 could interfere with various host cell functions, and to identify drug targets amongst the host genes that could potentially be modulated against COVID-19 by repurposing existing drugs. The machine learning model employed here involves gene embeddings that leverage causal gene expression signatures curated from literature. In contrast to other network-based approaches for drug repurposing, our approach explicitly takes the direction of effects into account, distinguishing between activation and inhibition.

RESULTS : We have constructed 70 networks connecting SARS-CoV-2 viral proteins to various biological functions, diseases, and pathways reflecting viral biology, clinical observations, and co-morbidities in the context of COVID-19. Results are presented in the form of interactive network visualizations through a web interface, the Coronavirus Network Explorer (CNE), that allows exploration of underlying experimental evidence. We find that existing drugs targeting genes in those networks are strongly enriched in the set of drugs that are already in clinical trials against COVID-19.

CONCLUSIONS : The approach presented here can identify biologically plausible hypotheses for COVID-19 pathogenesis, explicitly connected to the immunological, virological and pathological observations seen in SARS-CoV-2 infected patients. The discovery of repurposable drugs is driven by prior knowledge of relevant functional endpoints that reflect known viral biology or clinical observations, therefore suggesting potential mechanisms of action. We believe that the CNE offers relevant insights that go beyond more conventional network approaches, and can be a valuable tool for drug repurposing. The CNE is available at https://digitalinsights.qiagen.com/coronavirus-network-explorer .

Krämer Andreas, Billaud Jean-Noël, Tugendreich Stuart, Shiffman Dan, Jones Martin, Green Jeff

2021-May-03

COVID-19, Drug repurposing, Knowledge graph, Network biology

General General

Change Matters: Medication Change Prediction with Recurrent Residual Networks

ArXiv Preprint

Deep learning is revolutionizing predictive healthcare, including recommending medications to patients with complex health conditions. Existing approaches focus on predicting all medications for the current visit, which often overlaps with medications from previous visits. A more clinically relevant task is to identify medication changes. In this paper, we propose a new recurrent residual network, named MICRON, for medication change prediction. MICRON takes the changes in patient health records as input and learns to update a hidden medication vector and the medication set recurrently with a reconstruction design. The medication vector is like the memory cell that encodes longitudinal information of medications. Unlike traditional methods that require the entire patient history for prediction, MICRON has a residual-based inference that allows for sequential updating based only on new patient features (e.g., new diagnoses in the recent visit) more efficiently. We evaluated MICRON on real inpatient and outpatient datasets. MICRON achieves 3.5% and 7.8% relative improvements over the best baseline in F1 score, respectively. MICRON also requires fewer parameters, which significantly reduces the training time to 38.3s per epoch with 1.5x speed-up.

Chaoqi Yang, Cao Xiao, Lucas Glass, Jimeng Sun

2021-05-05

General General

Benchmarking deep learning splice prediction tools using functional splice assays.

In Human mutation ; h5-index 53.0

Hereditary disorders are frequently caused by genetic variants that affect pre-mRNA splicing. Whilst genetic variants in the canonical splice motifs are almost always disrupting splicing, the pathogenicity of variants in the non-canonical splice sites (NCSS) and deep intronic (DI) regions are difficult to predict. Multiple splice prediction tools have been developed for this purpose, with the latest tools employing deep learning algorithms. We benchmarked established and deep learning splice prediction tools on published gold standard sets of 71 NCSS and 81 DI variants in the ABCA4 gene and 61 NCSS variants in the MYBPC3 gene with functional assessment in midigene and minigene splice assays. The selection of splice prediction tools included CADD, DSSP, GeneSplicer, MaxEntScan, MMSplice, MTSplice, NNSPLICE, SPIDEX, SpliceAI, SpliceRover and SpliceSiteFinder-like. The best performing splice prediction tool for the different variants was SpliceRover for ABCA4 NCSS variants, SpliceAI for ABCA4 DI variants and SpliceSiteFinder-like for NCSS variants in MYBPC3. Overall, the performance in a real time clinical setting is much more modest than reported by the developers of the tools. This article is protected by copyright. All rights reserved.

Riepe Tabea V, Khan Mubeen, Roosing Susanne, Cremers Frans P M, ‘t Hoen Peter A C

2021-May-03

ABCA4, MYBPC3, RNA splicing, deep learning, splice prediction tools, variant effect prediction

Public Health Public Health

The Translational Machine: A novel machine-learning approach to illuminate complex genetic architectures.

In Genetic epidemiology

The Translational Machine (TM) is a machine learning (ML)-based analytic pipeline that translates genotypic/variant call data into biologically contextualized features that richly characterize complex variant architectures and permit greater interpretability and biological replication. It also reduces potentially confounding effects of population substructure on outcome prediction. The TM consists of three main components. First, replicable but flexible feature engineering procedures translate genome-scale data into biologically informative features that appropriately contextualize simple variant calls/genotypes within biological and functional contexts. Second, model-free, nonparametric ML-based feature filtering procedures empirically reduce dimensionality and noise of both original genotype calls and engineered features. Third, a powerful ML algorithm for feature selection is used to differentiate risk variant contributions across variant frequency and functional prediction spectra. The TM simultaneously evaluates potential contributions of variants operative under polygenic and heterogeneous models of genetic architecture. Our TM enables integration of biological information (e.g., genomic annotations) within conceptual frameworks akin to geneset-/pathways-based and collapsing methods, but overcomes some of these methods' limitations. The full TM pipeline is executed in R. Our approach and initial findings from its application to a whole-exome schizophrenia case-control data set are presented. These TM procedures extend the findings of the primary investigation and yield novel results.

Askland Kathleen D, Strong David, Wright Marvin N, Moore Jason H

2021-May-03

genetic architecture, genetic heterogeneity, machine learning, psychiatric genetics, random forests, rare variants

Radiology Radiology

Automatic prediction of left cardiac chamber enlargement from chest radiographs using convolutional neural network.

In European radiology ; h5-index 62.0

OBJECTIVE : To develop deep learning-based cardiac chamber enlargement-detection algorithms for left atrial (DLCE-LAE) and ventricular enlargement (DLCE-LVE), on chest radiographs METHODS: For training and internal validation of DLCE-LAE and -LVE, 5,045 chest radiographs (CRs; 2,463 normal and 2,393 LAE) and 1,012 CRs (456 normal and 456 LVE) matched with the same-day echocardiography were collected, respectively. External validation was performed using 107 temporally independent CRs. Reader performance test was conducted using the external validation dataset by five cardiothoracic radiologists without and with the results of DLCE. Classification performance of DLCE was evaluated and compared with those of the readers and conventional radiographic features, including cardiothoracic ratio, carinal angle, and double contour. In addition, DLCE-LAE was tested on 5,277 CRs from a healthcare screening program cohort.

RESULTS : DLCE-LAE showed areas under the receiver operating characteristics curve (AUROCs) of 0.858 on external validation. On reader performance test, DLCE-LAE showed better results than pooled radiologists (AUROC 0.858 vs. 0.651; p < .001) and significantly increased their performance when used as a second reader (AUROC 0.651 vs. 0.722; p < .001). DLCE-LAE also showed a significantly higher AUROC than conventional radiographic findings (AUROC 0.858 vs. 0.535-0.706; all ps < .01). In the healthcare screening cohort, DLCE-LAE successfully detected 71.0% (142/200) CRs with moderate-to-severe LAE (93.5% [29/31] of severe cases), while yielding 11.8% (492/4,184) false-positive rate. DLCE-LVE showed AUROCs of 0.966 and 0.594 on internal and external validation, respectively.

CONCLUSION : DLCE-LAE outperformed and improved cardiothoracic radiologists' performance in detecting LAE and showed promise in screening individuals with moderate-to-severe LAE in a healthcare screening cohort.

KEY POINTS : • Our deep learning algorithm outperformed cardiothoracic radiologists in detecting left atrial enlargement on chest radiographs. • Cardiothoracic radiologists improved their performance in detecting left atrial enlargement when aided by the algorithm. • On a healthcare-screening cohort, our algorithm detected 71.0% (142/200) radiographs with moderate-to-severe left atrial enlargement while yielding 11.8% (492/4,184) false-positive rate.

Nam Ju Gang, Kim Jinwook, Noh Keonwoo, Choi Hyewon, Kim Da Som, Yoo Seung-Jin, Yang Hyun-Lim, Hwang Eui Jin, Goo Jin Mo, Park Eun-Ah, Sun Hye Young, Kim Min-Soo, Park Chang Min

2021-May-03

Diagnosis, computer-assisted, Left atrium, X-ray film

General General

Investigating Factors of Active Aging among Chinese Older Adults: A Machine Learning Approach.

In The Gerontologist

BACKGROUND AND OBJECTIVES : With the extension of healthy life expectancy, promoting active aging has become a policy response to rapid population aging in China. Yet, it has been inconclusive about the relative importance of the determinants of active aging. By applying a machine learning approach, this study aims to identify the most important determinants of active aging in three domains, i.e., paid/unpaid work, caregiving, and social activities, among Chinese older adults.

RESEARCH DESIGN AND METHOD : Data were drawn from the first wave of the China Health and Retirement Longitudinal Study (CHARLS), which surveys a nationally representative sample of adults aged 60-year-old and above (N=7,503). We estimated Random Forest and the least absolute shrinkage and selection operator (LASSO) regression models to determine the most important factors related to active aging.

RESULTS : Health has a generic effect on all outcomes of active aging. Our findings also identified the domain-specific determinants of active aging. Urban/rural residency is among the most important factors determining the likelihood of engaging in paid/unpaid work. Living in a multi-generational household is especially important in predicting caregiving activities. Neighborhood infrastructure and facilities have the strongest influence on older adults' participation in social activities.

DISCUSSION AND IMPLICATIONS : The application of feature selection models provides a fruitful first step in identifying the most important determinants of active aging among Chinese older adults. These results provide evidence-based recommendations for policies and practices promoting active aging.

Yu Jiao, Huang Wenxuan, Kahana Eva

2021-May-04

Active aging, Chinese context, LASSO regression, Random Forest

General General

News Sentiment Informed Time-series Analyzing AI (SITALA) to curb the spread of COVID-19 in Houston.

In Expert systems with applications

Coronavirus disease (COVID-19) has evolved into a pandemic with many unknowns. Houston, located in the Harris County of Texas, is becoming the next hotspot of this pandemic. With a severe decline in international and inter-state travel, a model at the county level is needed as opposed to the state or country level. Existing approaches have a few drawbacks. Firstly, the data used is the number of COVID-19 positive cases instead of positivity. The former is a function of the number of tests carried out while the number of tests normalizes the latter. Positivity gives a better picture of the spread of this pandemic as, with time, more tests are being administered. Positivity under 5% has been desired for the reopening of businesses to almost 100% capacity. Secondly, the data used by models like SEIRD (Susceptible, Exposed, Infectious, Recovered, and Deceased) lacks information about the sentiment of people concerning coronavirus. Thirdly, models that make use of social media posts might have too much noise and misinformation. On the other hand, news sentiment can capture long-term effects of hidden variables like public policy, opinions of local doctors, and disobedience of state-wide mandates. The present study introduces a new artificial intelligence (i.e., AI) model, viz., Sentiment Informed Time-series Analyzing AI (SITALA), trained on COVID-19 test positivity data and news sentiment from over 2750 news articles for Harris county. The news sentiment was obtained using IBM Watson Discovery News. SITALA is inspired by Google-Wavenet architecture and makes use of TensorFlow. The mean absolute error for the training dataset of 66 consecutive days is 2.76, and that for the test dataset of 22 consecutive days is 9.6. A cone of uncertainty is provided within which future COVID-19 test positivity has been shown to fall with high accuracy. The model predictions fare better than a published Bayesian-based SEIRD model. The model forecasts that in order to curb the spread of coronavirus in Houston, a sustained negative news sentiment (e.g., death count for COVID-19 will grow at an alarming rate in Houston if mask orders are not followed) will be desirable. Public policymakers may use SITALA to set the tone of the local policies and mandates.

Desai Prathamesh S

2021-Apr-29

Artificial Intelligence, COVID-19 Model, Deep Learning, News Sentiment, Pandemic Forecast, Public Policy

Public Health Public Health

Multimodal deep learning from satellite and street-level imagery for measuring income, overcrowding, and environmental deprivation in urban areas.

In Remote sensing of environment

Data collected at large scale and low cost (e.g. satellite and street level imagery) have the potential to substantially improve resolution, spatial coverage, and temporal frequency of measurement of urban inequalities. Multiple types of data from different sources are often available for a given geographic area. Yet, most studies utilize a single type of input data when making measurements due to methodological difficulties in their joint use. We propose two deep learning-based methods for jointly utilizing satellite and street level imagery for measuring urban inequalities. We use London as a case study for three selected outputs, each measured in decile classes: income, overcrowding, and environmental deprivation. We compare the performances of our proposed multimodal models to corresponding unimodal ones using mean absolute error (MAE). First, satellite tiles are appended to street level imagery to enhance predictions at locations where street images are available leading to improvements in accuracy by 20, 10, and 9% in units of decile classes for income, overcrowding, and living environment. The second approach, novel to the best of our knowledge, uses a U-Net architecture to make predictions for all grid cells in a city at high spatial resolution (e.g. for 3 m × 3 m pixels in London in our experiments). It can utilize city wide availability of satellite images as well as more sparse information from street-level images where they are available leading to improvements in accuracy by 6, 10, and 11%. We also show examples of prediction maps from both approaches to visually highlight performance differences.

Suel Esra, Bhatt Samir, Brauer Michael, Flaxman Seth, Ezzati Majid

2021-May

Convolutional neural networks, Satellite images, Segmentation, Street-level images, Urban measurements

oncology Oncology

Evolution of delayed resistance to immunotherapy in a melanoma responder.

In Nature medicine ; h5-index 170.0

Despite initial responses1-3, most melanoma patients develop resistance4 to immune checkpoint blockade (ICB). To understand the evolution of resistance, we studied 37 tumor samples over 9 years from a patient with metastatic melanoma with complete clinical response to ICB followed by delayed recurrence and death. Phylogenetic analysis revealed co-evolution of seven lineages with multiple convergent, but independent resistance-associated alterations. All recurrent tumors emerged from a lineage characterized by loss of chromosome 15q, with post-treatment clones acquiring additional genomic driver events. Deconvolution of bulk RNA sequencing and highly multiplexed immunofluorescence (t-CyCIF) revealed differences in immune composition among different lineages. Imaging revealed a vasculogenic mimicry phenotype in NGFRhi tumor cells with high PD-L1 expression in close proximity to immune cells. Rapid autopsy demonstrated two distinct NGFR spatial patterns with high polarity and proximity to immune cells in subcutaneous tumors versus a diffuse spatial pattern in lung tumors, suggesting different roles of this neural-crest-like program in different tumor microenvironments. Broadly, this study establishes a high-resolution map of the evolutionary dynamics of resistance to ICB, characterizes a de-differentiated neural-crest tumor population in melanoma immunotherapy resistance and describes site-specific differences in tumor-immune interactions via longitudinal analysis of a patient with melanoma with an unusual clinical course.

Liu David, Lin Jia-Ren, Robitschek Emily J, Kasumova Gyulnara G, Heyde Alex, Shi Alvin, Kraya Adam, Zhang Gao, Moll Tabea, Frederick Dennie T, Chen Yu-An, Wang Shu, Schapiro Denis, Ho Li-Lun, Bi Kevin, Sahu Avinash, Mei Shaolin, Miao Benchun, Sharova Tatyana, Alvarez-Breckenridge Christopher, Stocking Jackson H, Kim Tommy, Fadden Riley, Lawrence Donald, Hoang Mai P, Cahill Daniel P, Malehmir Mohsen, Nowak Martin A, Brastianos Priscilla K, Lian Christine G, Ruppin Eytan, Izar Benjamin, Herlyn Meenhard, Van Allen Eliezer M, Nathanson Katherine, Flaherty Keith T, Sullivan Ryan J, Kellis Manolis, Sorger Peter K, Boland Genevieve M

2021-May-03

General General

Machine learning applied to near-infrared spectra for clinical pleural effusion classification.

In Scientific reports ; h5-index 158.0

Lung cancer patients with malignant pleural effusions (MPE) have a particular poor prognosis. It is crucial to distinguish MPE from benign pleural effusion (BPE). The present study aims to develop a rapid, convenient and economical diagnostic method based on FTIR near-infrared spectroscopy (NIRS) combined with machine learning strategy for clinical pleural effusion classification. NIRS spectra were recorded for 47 MPE samples and 35 BPE samples. The sample data were randomly divided into train set (n = 62) and test set (n = 20). Partial least squares, random forest, support vector machine (SVM), and gradient boosting machine models were trained, and subsequent predictive performance were predicted on the test set. Besides the whole spectra used in modeling, selected features using SVM recursive feature elimination algorithm were also investigated in modeling. Among those models, NIRS combined with SVM showed the best predictive performance (accuracy: 1.0, kappa: 1.0, and AUCROC: 1.0). SVM with the top 50 feature wavenumbers also displayed a high predictive performance (accuracy: 0.95, kappa: 0.89, AUCROC: 0.99). Our study revealed that the combination of NIRS and machine learning is an innovative, rapid, and convenient method for clinical pleural effusion classification, and worth further evaluation.

Chen Zhongjian, Chen Keke, Lou Yan, Zhu Jing, Mao Weimin, Song Zhengbo

2021-May-03

General General

Belief polarization in a complex world: A learning theory perspective.

In Proceedings of the National Academy of Sciences of the United States of America

We present two models of how people form beliefs that are based on machine learning theory. We illustrate how these models give insight into observed human phenomena by showing how polarized beliefs can arise even when people are exposed to almost identical sources of information. In our first model, people form beliefs that are deterministic functions that best fit their past data (training sets). In that model, their inability to form probabilistic beliefs can lead people to have opposing views even if their data are drawn from distributions that only slightly disagree. In the second model, people pay a cost that is increasing in the complexity of the function that represents their beliefs. In this second model, even with large training sets drawn from exactly the same distribution, agents can disagree substantially because they simplify the world along different dimensions. We discuss what these models of belief formation suggest for improving people's accuracy and agreement.

Haghtalab Nika, Jackson Matthew O, Procaccia Ariel D

2021-May-11

belief polarization, learning theory

Surgery Surgery

Surgical training fit for the future: the need for a change.

In Postgraduate medical journal

Postgraduate training in surgical specialties is one of the longest training programmes in the medical field. Most of the surgical training programmes require 5-6 years of postgraduate training to become qualified. This is usually followed by 1-2 years of fellowship training in a subspecialised interest. This has been the case for the last 20-30 years with no significant change. The surgical practice is transforming quickly due to the advances in medical technology. This transformation is not matched in the postgraduate training, there is minimal exposure to the new technological advances in early years of postgraduate training. The current postgraduate training in surgical specialties is not fit for the future. Early exposure to robotic and artificial intelligence technologies is required. To achieve this, a significant transformation of surgical training is necessary, which requires a new vision and involves significant investment. We discuss the need for this transformation in the postgraduate surgical specialties training and analyse the threats and opportunities in relation to this transformation.

Elnikety Sherif, Badr Eman, Abdelaal Ahmed

2021-May-03

medical education & training, surgery

General General

Designing roadside green infrastructure to mitigate traffic-related air pollution using machine learning.

In The Science of the total environment

Communities located in near-road environments are exposed to traffic-related air pollution (TRAP), causing adverse health effects. While roadside vegetation barriers can help mitigate TRAP, their effectiveness to reduce TRAP is influenced by site-specific conditions. To test vegetation designs using direct field measurements or high-fidelity numerical simulations is often infeasible since urban planners and local communities often lack the access and expertise to use those tools. There is a need for a fast, reliable, and easy-to-use method to evaluate vegetation barrier designs based on their capacity to mitigate TRAP. In this paper, we investigated five machine learning (ML) methods, including linear regression (LR), support vector machine (SVM), random forest (RF), XGBoost (XGB), and neural networks (NN), to predict size-resolved and locationally dependent particle concentrations downwind of various vegetation barrier designs. Data from 83 computational fluid dynamics (CFD) simulations was used to train and test the ML models. We developed downwind region-specific models to capture the complexity of this problem and enhance the overall accuracy. Our feature space was composed of variables that can be feasibly obtained such as vegetation width, height, leaf area index (LAI), particle size, leaf area density (LAD) and wind speed at different heights. RF, NN, and XGB performed well with a normalized root mean square error (NRMSE) of 6-7% and an average test R2 value >0.91, while SVM and LR had an NRMSE of approximately 13% and an average test R2 value of 0.56. Using feature selection, vegetation dimensions and particle size had the highest influence in predicting pollutant concentrations. The ML models developed can help create tools to aid local communities in developing mitigation strategies to address TRAP problems.

Hashad Khaled, Gu Jiajun, Yang Bo, Rong Morena, Chen Edric, Ma Xiaoxin, Zhang K Max

2021-Jun-15

Air quality, Green infrastructure, Machine learning, Particulate matter (PM), Urban designs, Vegetation

Public Health Public Health

Using machine learning to develop a novel COVID-19 Vulnerability Index (C19VI).

In The Science of the total environment

COVID-19 is now one of the most leading causes of death in the United States (US). Systemic health, social and economic disparities have put the minorities and economically poor communities at a higher risk than others. There is an immediate requirement to develop a reliable measure of county-level vulnerabilities that can capture the heterogeneity of vulnerable communities. This study reports a COVID-19 Vulnerability Index (C19VI) for identifying and mapping vulnerable counties. We proposed a Random Forest machine learning-based vulnerability model using CDC's sociodemographic and COVID-19-specific themes. An innovative 'COVID-19 Impact Assessment' algorithm was also developed for evaluating severity of the pandemic and to train the vulnerability model. Developed C19VI was statistically validated and compared with the CDC COVID-19 Community Vulnerability Index (CCVI). Finally, using C19VI and the census data, we explored racial inequalities and economic disparities in COVID-19 health outcomes. Our index indicates that 575 counties (45 million people) fall into the 'very high' vulnerability class, 765 counties (66 million people) in the 'high' vulnerability class, and 1435 counties (204 million people) in the 'moderate' or 'low' vulnerability class. Only 367 counties (20 million people) were found as 'very low' vulnerable areas. Furthermore, C19VI reveals that 524 counties with a racial minority population higher than 13% and 420 counties with poverty higher than 20% are in the 'very high' or 'high' vulnerability classes. The C19VI aims at helping public health officials and disaster management agencies to develop effective mitigation strategies especially for the disproportionately impacted communities.

Tiwari Anuj, Dadhania Arya V, Ragunathrao Vijay Avin Balaji, Oliveira Edson R A

2021-Jun-15

COVID-19, Disproportionate COVID-19, Machine learning, Racial minority, Vulnerability modeling

oncology Oncology

Lesion Segmentation and RECIST Diameter Prediction via Click-driven Attention and Dual-path Connection

ArXiv Preprint

Measuring lesion size is an important step to assess tumor growth and monitor disease progression and therapy response in oncology image analysis. Although it is tedious and highly time-consuming, radiologists have to work on this task by using RECIST criteria (Response Evaluation Criteria In Solid Tumors) routinely and manually. Even though lesion segmentation may be the more accurate and clinically more valuable means, physicians can not manually segment lesions as now since much more heavy laboring will be required. In this paper, we present a prior-guided dual-path network (PDNet) to segment common types of lesions throughout the whole body and predict their RECIST diameters accurately and automatically. Similar to [1], a click guidance from radiologists is the only requirement. There are two key characteristics in PDNet: 1) Learning lesion-specific attention matrices in parallel from the click prior information by the proposed prior encoder, named click-driven attention; 2) Aggregating the extracted multi-scale features comprehensively by introducing top-down and bottom-up connections in the proposed decoder, named dual-path connection. Experiments show the superiority of our proposed PDNet in lesion segmentation and RECIST diameter prediction using the DeepLesion dataset and an external test set. PDNet learns comprehensive and representative deep image features for our tasks and produces more accurate results on both lesion segmentation and RECIST diameter prediction.

Youbao Tang, Ke Yan, Jinzheng Cai, Lingyun Huang, Guotong Xie, Jing Xiao, Jingjing Lu, Gigin Lin, Le Lu

2021-05-05

oncology Oncology

Computationally Derived Cribriform Area Index from Prostate Cancer Hematoxylin and Eosin Images Is Associated with Biochemical Recurrence Following Radical Prostatectomy and Is Most Prognostic in Gleason Grade Group 2.

In European urology focus

BACKGROUND : The presence of invasive cribriform adenocarcinoma (ICC), an expanse of cells containing punched-out lumina uninterrupted by stroma, in radical prostatectomy (RP) specimens has been associated with biochemical recurrence (BCR). However, ICC identification has only moderate inter-reviewer agreement.

OBJECTIVE : To investigate quantitative machine-based assessment of the extent and prognostic utility of ICC, especially within individual Gleason grade groups.

DESIGN, SETTING, AND PARTICIPANTS : A machine learning approach was developed for ICC segmentation using 70 RP patients and validated in a cohort of 749 patients from four sites whose median year of surgery was 2007 and with median follow-up of 28 mo. ICC was segmented on one representative hematoxylin and eosin RP slide per patient and the fraction of tumor area composed of ICC, the cribriform area index (CAI), was measured.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS : The association between CAI and BCR was measured in terms of the concordance index (c index) and hazard ratio (HR).

RESULTS AND LIMITATIONS : CAI was correlated with BCR (c index 0.62) in the validation set of 411 patients with ICC morphology, especially those with Gleason grade group 2 cancer (n = 192; c index 0.66), and was less prognostic when patients without ICC were included (c index 0.54). A doubling of CAI in the group with ICC morphology was prognostic after controlling for Gleason grade, surgical margin positivity, preoperative prostate-specific antigen level, pathological T stage, and age (HR 1.19, 95% confidence interval 1.03-1.38; p = 0.018).

CONCLUSIONS : Automated image analysis and machine learning could provide an objective, quantitative, reproducible, and high-throughput method of quantifying ICC area. The performance of CAI for grade group 2 cancer suggests that for patients with little Gleason 4 pattern, the ICC fraction has a strong prognostic role.

PATIENT SUMMARY : Machine-based measurement of a specific cell pattern (cribriform; sieve-like, with lots of spaces) in images of prostate specimens could improve risk stratification for patients with prostate cancer. In the future, this could help in expanding the criteria for active surveillance.

Leo Patrick, Chandramouli Sacheth, Farré Xavier, Elliott Robin, Janowczyk Andrew, Bera Kaustav, Fu Pingfu, Janaki Nafiseh, El-Fahmawi Ayah, Shahait Mohammed, Kim Jessica, Lee David, Yamoah Kosj, Rebbeck Timothy R, Khani Francesca, Robinson Brian D, Shih Natalie N C, Feldman Michael, Gupta Sanjay, McKenney Jesse, Lal Priti, Madabhushi Anant

2021-Apr-30

Biochemical recurrence, Cribriform, Digital pathology, Gleason grading, Machine learning, Prostate cancer

Cardiology Cardiology

Ethical implications of AI in robotic surgical training: A Delphi consensus statement.

In European urology focus

CONTEXT : As the role of AI in healthcare continues to expand there is increasing awareness of the potential pitfalls of AI and the need for guidance to avoid them.

OBJECTIVES : To provide ethical guidance on developing narrow AI applications for surgical training curricula. We define standardised approaches to developing AI driven applications in surgical training that address current recognised ethical implications of utilising AI on surgical data. We aim to describe an ethical approach based on the current evidence, understanding of AI and available technologies, by seeking consensus from an expert committee.

EVIDENCE ACQUISITION : The project was carried out in 3 phases: (1) A steering group was formed to review the literature and summarize current evidence. (2) A larger expert panel convened and discussed the ethical implications of AI application based on the current evidence. A survey was created, with input from panel members. (3) Thirdly, panel-based consensus findings were determined using an online Delphi process to formulate guidance. 30 experts in AI implementation and/or training including clinicians, academics and industry contributed. The Delphi process underwent 3 rounds. Additions to the second and third-round surveys were formulated based on the answers and comments from previous rounds. Consensus opinion was defined as ≥ 80% agreement.

EVIDENCE SYNTHESIS : There was 100% response from all 3 rounds. The resulting formulated guidance showed good internal consistency, with a Cronbach alpha of >0.8. There was 100% consensus that there is currently a lack of guidance on the utilisation of AI in the setting of robotic surgical training. Consensus was reached in multiple areas, including: 1. Data protection and privacy; 2. Reproducibility and transparency; 3. Predictive analytics; 4. Inherent biases; 5. Areas of training most likely to benefit from AI.

CONCLUSIONS : Using the Delphi methodology, we achieved international consensus among experts to develop and reach content validation for guidance on ethical implications of AI in surgical training. Providing an ethical foundation for launching narrow AI applications in surgical training. This guidance will require further validation.

PATIENT SUMMARY : As the role of AI in healthcare continues to expand there is increasing awareness of the potential pitfalls of AI and the need for guidance to avoid them.In this paper we provide guidance on ethical implications of AI in surgical training.

Collins Justin W, Marcus Hani J, Ghazi Ahmed, Sridhar Ashwin, Hashimoto Daniel, Hager Gregory, Arezzo Alberto, Jannin Pierre, Maier-Hein Lena, Marz Keno, Valdastri Pietro, Mori Kensaku, Elson Daniel, Giannarou Stamatia, Slack Mark, Hares Luke, Beaulieu Yanick, Levy Jeff, Laplante Guy, Ramadorai Arvind, Jarc Anthony, Andrews Ben, Garcia Pablo, Neemuchwala Huzefa, Andrusaite Alina, Kimpe Tom, Hawkes David, Kelly John D, Stoyanov Danail

2021-Apr-30

Artificial intelligence, Computer vision, Deep learning, GDPR, Learning algorithms, Natural language processing, biases, curriculum development, data protection, machine learning, narrow AI, predictive analytics, privacy, risk prediction, surgical education, training, transparency

Surgery Surgery

Three-dimensional virtual planning in mandibular advancement surgery: Soft tissue prediction based on deep learning.

In Journal of cranio-maxillo-facial surgery : official publication of the European Association for Cranio-Maxillo-Facial Surgery

The study aimed at developing a deep-learning (DL)-based algorithm to predict the virtual soft tissue profile after mandibular advancement surgery, and to compare its accuracy with the mass tensor model (MTM). Subjects who underwent mandibular advancement surgery were enrolled and divided into a training group and a test group. The DL model was trained using 3D photographs and CBCT data based on surgically achieved mandibular displacements (training group). Soft tissue simulations generated by DL and MTM based on the actual surgical jaw movements (test group) were compared with soft-tissue profiles on postoperative 3D photographs using distance mapping in terms of mean absolute error in the lower face, lower lip, and chin regions. 133 subjects were included - 119 in the training group and 14 in the test group. The mean absolute error for DL-based simulations of the lower face region was 1.0 ± 0.6 mm and was significantly lower (p = 0.02) compared with MTM-based simulations (1.5 ± 0.5 mm). CONCLUSION: The DL-based algorithm can predict 3D soft tissue profiles following mandibular advancement surgery. With a clinically acceptable mean absolute error. Therefore, it seems to be a relevant option for soft tissue prediction in orthognathic surgery. Therefore, it seems to be a relevant options.

Ter Horst Rutger, van Weert Hanneke, Loonen Tom, Bergé Stefaan, Vinayahalingam Shank, Baan Frank, Maal Thomas, de Jong Guido, Xi Tong

2021-Apr-21

3D face analysis, Artificial intelligence, Deep learning, Mandibular advancement surgery, Mass tensor model, Orthognathic surgery, Soft tissue prediction

General General

Mechanisms for handling nested dependencies in neural-network language models and humans.

In Cognition

Recursive processing in sentence comprehension is considered a hallmark of human linguistic abilities. However, its underlying neural mechanisms remain largely unknown. We studied whether a modern artificial neural network trained with "deep learning" methods mimics a central aspect of human sentence processing, namely the storing of grammatical number and gender information in working memory and its use in long-distance agreement (e.g., capturing the correct number agreement between subject and verb when they are separated by other phrases). Although the network, a recurrent architecture with Long Short-Term Memory units, was solely trained to predict the next word in a large corpus, analysis showed the emergence of a very sparse set of specialized units that successfully handled local and long-distance syntactic agreement for grammatical number. However, the simulations also showed that this mechanism does not support full recursion and fails with some long-range embedded dependencies. We tested the model's predictions in a behavioral experiment where humans detected violations in number agreement in sentences with systematic variations in the singular/plural status of multiple nouns, with or without embedding. Human and model error patterns were remarkably similar, showing that the model echoes various effects observed in human data. However, a key difference was that, with embedded long-range dependencies, humans remained above chance level, while the model's systematic errors brought it below chance. Overall, our study shows that exploring the ways in which modern artificial neural networks process sentences leads to precise and testable hypotheses about human linguistic performance.

Lakretz Yair, Hupkes Dieuwke, Vergallito Alessandra, Marelli Marco, Baroni Marco, Dehaene Stanislas

2021-Apr-30

Grammatical agreement, Language models, Long-range dependencies, Recurrent neural networks, Recursion, Relative clauses, Syntactic processing

Radiology Radiology

Machine learning-based radiomic evaluation of treatment response prediction in glioblastoma.

In Clinical radiology

AIM : To investigate machine learning based models combining clinical, radiomic, and molecular information to distinguish between early true progression (tPD) and pseudoprogression (psPD) in patients with glioblastoma.

MATERIALS AND METHODS : A retrospective analysis was undertaken of 76 patients (46 tPD, 30 psPD) with early enhancing disease following chemoradiotherapy for glioblastoma. Outcome was determined on follow-up until 6 months post-chemoradiotherapy. Models comprised clinical characteristics, O6-methylguanine-DNA methyltransferase (MGMT) promoter methylation status, and 307 quantitative imaging features extracted from enhancing disease and perilesional oedema masks on early post-chemoradiotherapy contrast-enhanced T1-weighted imaging, T2-weighted imaging (T2WI), and apparent diffusion coefficient (ADC) maps. Feature selection was performed within bootstrapped cross-validated recursive feature elimination with a random forest algorithm. Naive Bayes five-fold cross-validation was used to validate the final model.

RESULTS : Top selected features included age, MGMT promoter methylation status, two shape-based features from the enhancing disease mask, three radiomic features from the enhancing disease mask on ADC, and one radiomic feature from the perilesional oedema mask on T2WI. The final model had an area under the receiver operating characteristics curve (AUC) of 0.80, sensitivity 78.2%, specificity 66.7%, and accuracy of 73.7%.

CONCLUSION : Incorporating a machine learning-based approach using quantitative radiomic features from standard-of-care magnetic resonance imaging (MRI), in combination with clinical characteristics and MGMT promoter methylation status has a complementary effect and improves model performance for early prediction of glioblastoma treatment response.

Patel M, Zhan J, Natarajan K, Flintham R, Davies N, Sanghera P, Grist J, Duddalwar V, Peet A, Sawlani V

2021-Apr-30

General General

Artificial intelligence-based computational framework for drug-target prioritization and inference of novel repositionable drugs for Alzheimer's disease.

In Alzheimer's research & therapy ; h5-index 49.0

BACKGROUND : Identifying novel therapeutic targets is crucial for the successful development of drugs. However, the cost to experimentally identify therapeutic targets is huge and only approximately 400 genes are targets for FDA-approved drugs. As a result, it is inevitable to develop powerful computational tools that can identify potential novel therapeutic targets. Fortunately, the human protein-protein interaction network (PIN) could be a useful resource to achieve this objective.

METHODS : In this study, we developed a deep learning-based computational framework that extracts low-dimensional representations of high-dimensional PIN data. Our computational framework uses latent features and state-of-the-art machine learning techniques to infer potential drug target genes.

RESULTS : We applied our computational framework to prioritize novel putative target genes for Alzheimer's disease and successfully identified key genes that may serve as novel therapeutic targets (e.g., DLG4, EGFR, RAC1, SYK, PTK2B, SOCS1). Furthermore, based on these putative targets, we could infer repositionable candidate-compounds for the disease (e.g., tamoxifen, bosutinib, and dasatinib).

CONCLUSIONS : Our deep learning-based computational framework could be a powerful tool to efficiently prioritize new therapeutic targets and enhance the drug repositioning strategy.

Tsuji Shingo, Hase Takeshi, Yachie-Kinoshita Ayako, Nishino Taiko, Ghosh Samik, Kikuchi Masataka, Shimokawa Kazuro, Aburatani Hiroyuki, Kitano Hiroaki, Tanaka Hiroshi

2021-May-03

Deep learning, Drug discovery, Machine learning, Network embedding, Protein interaction network, Systems biology

General General

Schema: metric learning enables interpretable synthesis of heterogeneous single-cell modalities.

In Genome biology ; h5-index 114.0

A complete understanding of biological processes requires synthesizing information across heterogeneous modalities, such as age, disease status, or gene expression. Technological advances in single-cell profiling have enabled researchers to assay multiple modalities simultaneously. We present Schema, which uses a principled metric learning strategy that identifies informative features in a modality to synthesize disparate modalities into a single coherent interpretation. We use Schema to infer cell types by integrating gene expression and chromatin accessibility data; demonstrate informative data visualizations that synthesize multiple modalities; perform differential gene expression analysis in the context of spatial variability; and estimate evolutionary pressure on peptide sequences.

Singh Rohit, Hie Brian L, Narayan Ashwin, Berger Bonnie

2021-May-03

General General

A maximum flow-based network approach for identification of stable noncoding biomarkers associated with the multigenic neurological condition, autism.

In BioData mining

BACKGROUND : Machine learning approaches for predicting disease risk from high-dimensional whole genome sequence (WGS) data often result in unstable models that can be difficult to interpret, limiting the identification of putative sets of biomarkers. Here, we design and validate a graph-based methodology based on maximum flow, which leverages the presence of linkage disequilibrium (LD) to identify stable sets of variants associated with complex multigenic disorders.

RESULTS : We apply our method to a previously published logistic regression model trained to identify variants in simple repeat sequences associated with autism spectrum disorder (ASD); this L1-regularized model exhibits high predictive accuracy yet demonstrates great variability in the features selected from over 230,000 possible variants. In order to improve model stability, we extract the variants assigned non-zero weights in each of 5 cross-validation folds and then assemble the five sets of features into a flow network subject to LD constraints. The maximum flow formulation allowed us to identify 55 variants, which we show to be more stable than the features identified by the original classifier.

CONCLUSION : Our method allows for the creation of machine learning models that can identify predictive variants. Our results help pave the way towards biomarker-based diagnosis methods for complex genetic disorders.

Varma Maya, Paskov Kelley M, Chrisman Brianna S, Sun Min Woo, Jung Jae-Yoon, Stockham Nate T, Washington Peter Y, Wall Dennis P

2021-May-03

Feature selection, Feature stability, Linkage disequilibrium, Machine learning, Maximum flow, Network

General General

Leaf area index estimation model for UAV image hyperspectral data based on wavelength variable selection and machine learning methods.

In Plant methods

BACKGROUND : To accurately estimate winter wheat leaf area index (LAI) using unmanned aerial vehicle (UAV) hyperspectral imagery is crucial for crop growth monitoring, fertilization management, and development of precision agriculture.

METHODS : The UAV hyperspectral imaging data, Analytical Spectral Devices (ASD) data, and LAI were simultaneously obtained at main growth stages (jointing stage, booting stage, and filling stage) of various winter wheat varieties under various nitrogen fertilizer treatments. The characteristic bands related to LAI were extracted from UAV hyperspectral data with different algorithms including first derivative (FD), successive projections algorithm (SPA), competitive adaptive reweighed sampling (CARS), and competitive adaptive reweighed sampling combined with successive projections algorithm (CARS_SPA). Furthermore, three modeling machine learning methods including partial least squares regression (PLSR), support vector machine regression (SVR), and extreme gradient boosting (Xgboost) were used to build LAI estimation models.

RESULTS : The results show that the correlation coefficient between UAV and ASD hyperspectral data is greater than 0.99, indicating the UAV data can be used for estimation of wheat growth information. The LAI bands selected by using different algorithms were slightly different among the 15 models built in this study. The Xgboost model using nine consecutive characteristic bands selected by CARS_SPA algorithm as input was proved to have the best performance. This model yielded identical results of coefficient of determination (0.89) for both calibration set and validation set, indicating a high accuracy of this model.

CONCLUSIONS : The Xgboost modeling method in combine with CARS_SPA algorithm can reduce input variables and improve the efficiency of model operation. The results provide reference and technical support for nondestructive and rapid estimation of winter wheat LAI by using UAV.

Zhang Juanjuan, Cheng Tao, Guo Wei, Xu Xin, Qiao Hongbo, Xie Yimin, Ma Xinming

2021-May-03

Characteristic bands, Hyperspectral imaging data, Leaf area index, Machine learning, Model, Unmanned aerial vehicle, Winter wheat

Surgery Surgery

Artificial intelligence improves the accuracy of residents in the diagnosis of hip fractures: a multicenter study.

In BMC musculoskeletal disorders ; h5-index 46.0

BACKGROUND : Less experienced clinicians sometimes make misdiagnosis of hip fractures. We developed computer-aided diagnosis (CAD) system for hip fractures on plain X-rays using a deep learning model trained on a large dataset. In this study, we examined whether the accuracy of the diagnosis of hip fracture of the residents could be improved by using this system.

METHODS : A deep convolutional neural network approach was used for machine learning. Pytorch 1.3 and Fast.ai 1.0 were applied as frameworks, and an EfficientNet-B4 model (a pre-trained ImageNet model) was used. We handled the 5295 X-rays from the patients with femoral neck fracture or femoral trochanteric fracture from 2009 to 2019. We excluded cases in which the bilateral hips were not included within an image range, and cases of femoral shaft fracture and periprosthetic fracture. Finally, we included 5242 AP pelvic X-rays from 4851 cases. We divided these 5242 images into two images per image, and prepared 5242 images including fracture site and 5242 images without fracture site. Thus, a total of 10,484 images were used for machine learning. The accuracy, sensitivity, specificity, F-value, and area under the curve (AUC) were assessed. Gradient-weighted class activation mapping (Grad-CAM) was used to conceptualize the basis for the diagnosis of the fracture by the deep learning algorithm. Secondly, we conducted a controlled experiment with clinicians. Thirty-one residents;young doctors within 2 years of graduation from medical school who rotate through various specialties, were tested using 300 hip fracture images that were randomly extracted from the dataset. We evaluated the diagnostic accuracy with and without the use of the CAD system for each of the 300 images.

RESULTS : The accuracy, sensitivity, specificity, F-value, and AUC were 96.1, 95.2, 96.9%, 0.961, and 0.99, respectively, with the correct diagnostic basis generated by Grad-CAM. In the controlled experiment, the diagnostic accuracy of the residents significantly improved when they used the CAD system.

CONCLUSIONS : We developed a newly CAD system with a deep learning algorithm from a relatively large dataset from multiple institutions. Our system achieved high diagnostic performance. Our system improved the diagnostic accuracy of residents for hip fractures.

LEVEL OF EVIDENCE : Level III, Foundational evidence, before-after study.

CLINICAL RELEVANCE : high.

Sato Yoichi, Takegami Yasuhiko, Asamoto Takamune, Ono Yutaro, Hidetoshi Tsugeno, Goto Ryosuke, Kitamura Akira, Honda Seiwa

2021-May-03

General General

Service robots for affective labor: a sociology of labor perspective.

In AI & society

Profit-oriented service sectors such as tourism, hospitality, and entertainment are increasingly looking at how professional service robots can be integrated into the workplace to perform socio-cognitive tasks that were previously reserved for humans. This is a work in which social and labor sciences recognize the principle role of emotions. However, the models and narratives of emotions that drive research, design, and deployment of service robots in human-robot interaction differ considerably from how emotions are framed in the sociology of labor and feminist studies of service work. In this paper, we explore these tensions through the concepts of affective and emotional labor, and outline key insights these concepts offer for the design and evaluation of professional service robots. Taken together, an emphasis on interactionist approaches to emotions and on the demands of affective labor, leads us to argue that service employees are under-represented in existing studies in human-robot interaction. To address this, we outline how participatory design and value-sensitive design approaches can be applied as complimentary methodological frameworks that include service employees as vital stakeholders.

Dobrosovestnova Anna, Hannibal Glenda, Reinboth Tim

2021-Apr-28

Affective labor, Human–robot interaction, Service economy

General General

Perspective: Wearable Internet of Medical Things for Remote Tracking of Symptoms, Prediction of Health Anomalies, Implementation of Preventative Measures, and Control of Virus Spread During the Era of COVID-19.

In Frontiers in robotics and AI

The COVID-19 pandemic has highly impacted the communities globally by reprioritizing the means through which various societal sectors operate. Among these sectors, healthcare providers and medical workers have been impacted prominently due to the massive increase in demand for medical services under unprecedented circumstances. Hence, any tool that can help the compliance with social guidelines for COVID-19 spread prevention will have a positive impact on managing and controlling the virus outbreak and reducing the excessive burden on the healthcare system. This perspective article disseminates the perspectives of the authors regarding the use of novel biosensors and intelligent algorithms embodied in wearable IoMT frameworks for tackling this issue. We discuss how with the use of smart IoMT wearables certain biomarkers can be tracked for detection of COVID-19 in exposed individuals. We enumerate several machine learning algorithms which can be used to process a wide range of collected biomarkers for detecting (a) multiple symptoms of SARS-CoV-2 infection and (b) the dynamical likelihood of contracting the virus through interpersonal interaction. Eventually, we enunciate how a systematic use of smart wearable IoMT devices in various social sectors can intelligently help controlling the spread of COVID-19 in communities as they enter the reopening phase. We explain how this framework can benefit individuals and their medical correspondents by introducing Systems for Symptom Decoding (SSD), and how the use of this technology can be generalized on a societal level for the control of spread by introducing Systems for Spread Tracing (SST).

Mehrdad Sarmad, Wang Yao, Atashzar S Farokh

2021

AI for health, COVID-19, IoMT, smart connected health, smart wearables, spread control, symptom tracking, telemedicine

General General

Supporting the Early Detection of Disease Onset and Change Using Document Vector Analysis of Nursing Observation Records.

In Evaluation & the health professions

Nursing records are an account of patient condition and treatment during their hospital stay. In this study, we developed a system that can automatically analyze nursing records to predict the occurrence of diseases and incidents (e.g., falls). Text vectorization was performed for nursing records and compared with past case data on aspiration pneumonia, to develop an onset prediction system. Nursing records for a patient group that developed aspiration pneumonia during hospitalization and a non-onset control group were randomly assigned to definitive diagnostic (for learning), preliminary survey, and test datasets. Data from the preliminary survey were used to adjust parameters and influencing factors. The final verification used the test data and revealed the highest compatibility to predict the onset of aspiration pneumonia (sensitivity = 90.9%, specificity = 60.3%) with the parameter values of size = 80 (number of dimensions of the sentence vector), window = 13 (number of words before and after the learned word), and min_count = 2 (threshold of wordcount for word to be included). This method represents the foundation for a discovery/warning system using machine-based automated monitoring to predict the onset of diseases and prevent adverse incidents such as falls.

Komaki Shotaro, Muranaga Fuminori, Uto Yumiko, Iwaanakuchi Takashi, Kumamoto Ichiro

2021-May-03

aspiration pneumonia, machine learning, narrative medicine, natural language processing, patient safety

oncology Oncology

Single nucleotide polymorphism (SNP) array-based signature of low hypodiploidy in acute lymphoblastic leukemia.

In Genes, chromosomes & cancer

Low hypodiploidy (30-39 chromosomes) is one of the most prevalent genetic subtypes among adults with ALL and is associated with a very poor outcome. Low hypodiploid clones can often undergo a chromosomal doubling generating a near-triploid clone (60-78 chromosomes). When cytogenetic techniques detect a near triploid clone, a diagnostic challenge may ensue in differentiating presumed duplicated low hypodiploidy from good risk high hyperdiploid ALL (51-67 chromosomes). We used single-nucleotide polymorphism (SNP) arrays to analyze low hypodiploid/near triploid (HoTr) (n=48) and high hyperdiploid (HeH) (n=40) cases. In addition to standard analysis, we derived log2 ratios for entire chromosomes enabling us to analyze the cohort using machine-learning techniques. Low hypodiploid and near triploid cases clustered together and separately from high hyperdiploid samples. Using these approaches, we also identified three cases with 50-60 chromosomes, originally called as HeH, which were, in fact, HoTr and two cases incorrectly called as HoTr. TP53 mutation analysis supported the new classification of all cases tested. Next, we constructed a classification and regression tree model for predicting ploidy status with chromosomes 1, 7 and 14 being the key discriminators. The classifier correctly identified 47/50 (94%) HoTr cases. We validated the classifier using an independent cohort of 44 cases where it correctly called 7/7 (100%) low hypodiploid cases. The results of this study suggest that HoTr is more frequent among older adults with ALL than previously estimated and that SNP array analysis should accompany cytogenetics where possible. The classifier can assist where SNP array patterns are challenging to interpret. This article is protected by copyright. All rights reserved.

Creasey Thomas, Enshaei Amir, Nebral Karin, Schwab Claire, Watts Kathryn, Cuthbert Gavin, Vora Ajay, Moppett John, Harrison Christine J, Fielding Adele K, Haas Oskar A, Moorman Anthony V

2021-May-03

General General

Top predators govern multitrophic diversity effects in tritrophic food webs.

In Ecology

It is well known that functional diversity strongly affects ecosystem functioning. However, even in rather simple model communities consisting of only two or, at best, three trophic levels, the relationship between multitrophic functional diversity and ecosystem functioning appears difficult to generalize, due to its high contex-tuality. In this study, we considered several differently structured tritrophic food webs, in which the amount of functional diversity was varied independently on each trophic level. To achieve generalizable results, largely independent of parametriza-tion, we examined the outcomes of 128, 000 parameter combinations sampled from ecologically plausible intervals, with each tested for 200 randomly sampled initial conditions. Analysis of our data was done by training a Random Forest model. This method enables the identification of complex patterns in the data through partial dependence graphs, and the comparison of the relative influence of model parameters, including the degree of diversity, on food web properties. We found that bottom-up and top-down effects cascade simultaneously throughout the food web, intimately linking the effects of functional diversity of any trophic level to the amount of diversity of other trophic levels, which may explain the difficulty in unifying results from previous studies. Strikingly, only with high diversity throughout the whole food web, different interactions synergize to ensure efficient exploitation of the available nutrients and efficient biomass transfer to higher trophic levels, ultimately leading to a high biomass and production on the top level. The temporal variation of biomass showed a more complex pattern with increasing multitrophic diversity: while the system initially became less variable, eventually the temporal variation rose again due to the increasingly complex dynamical patterns. Importantly, top predator diversity and food web parameters affecting the top trophic level were of highest importance to determine the biomass and temporal variability of any trophic level. Overall, our study reveals that the mechanisms by which diversity influences ecosystem functioning are affected by every part of the food web, hampering the extrapolation of insights from simple monotrophic or bitrophic systems to complex natural food webs.

Ceulemans Ruben, Guill Christian, Gaedke Ursula

2021-May-03

food web efficiency, functional diversity, machine learning, nutrient exploitation, production, random forest, temporal variability, top predator, trait diversity

General General

Automated Detection of Enhanced DBS Device Settings.

In Companion Publication of the 2020 International Conference on Multimodal Interaction

Continuous deep brain stimulation (DBS) of the ventral striatum (VS) is an effective treatment for severe, treatment-refractory obsessive-compulsive disorder (OCD). Optimal parameter settings are signaled by a mirth response of intense positive affect, which are subjectively identified by clinicians. Subjective judgments are idiosyncratic and difficult to standardize. To objectively measure mirth responses, we used Automatic Facial Affect Recognition (AFAR) in a series of longitudinal assessments of a patient treated with DBS. Pre- and post-adjustment DBS were compared using both statistical and machine learning approaches. Positive affect was significantly higher post-DBS adjustment. Using SVM and XGBoost, participant's pre- and post-adjustment appearances were differentiated with F1 of 0.76, which suggests feasibility of objective measurement of mirth response.

Ding Yaohan, Ertugrul Itir Onal, Darzi Ali, Provenza Nicole, Jeni László A, Borton David, Goodman Wayne, Cohn Jeffrey

2020-Oct

DBS, OCD, affective computing, clinical research, ventral striatum

Public Health Public Health

Transcriptome prediction performance across machine learning models and diverse ancestries.

In HGG advances

Transcriptome prediction methods such as PrediXcan and FUSION have become popular in complex trait mapping. Most transcriptome prediction models have been trained in European populations using methods that make parametric linear assumptions like the elastic net (EN). To potentially further optimize imputation performance of gene expression across global populations, we built transcriptome prediction models using both linear and non-linear machine learning (ML) algorithms and evaluated their performance in comparison to EN. We trained models using genotype and blood monocyte transcriptome data from the Multi-Ethnic Study of Atherosclerosis (MESA) comprising individuals of African, Hispanic, and European ancestries and tested them using genotype and whole-blood transcriptome data from the Modeling the Epidemiology Transition Study (METS) comprising individuals of African ancestries. We show that the prediction performance is highest when the training and the testing population share similar ancestries regardless of the prediction algorithm used. While EN generally outperformed random forest (RF), support vector regression (SVR), and K nearest neighbor (KNN), we found that RF outperformed EN for some genes, particularly between disparate ancestries, suggesting potential robustness and reduced variability of RF imputation performance across global populations. When applied to a high-density lipoprotein (HDL) phenotype, we show including RF prediction models in PrediXcan revealed potential gene associations missed by EN models. Therefore, by integrating other ML modeling into PrediXcan and diversifying our training populations to include more global ancestries, we may uncover new genes associated with complex traits.

Okoro Paul C, Schubert Ryan, Guo Xiuqing, Johnson W Craig, Rotter Jerome I, Hoeschele Ina, Liu Yongmei, Im Hae Kyung, Luke Amy, Dugas Lara R, Wheeler Heather E

2021-Apr-08

General General

Deep learning and machine vision for food processing: A survey.

In Current research in food science

The quality and safety of food is an important issue to the whole society, since it is at the basis of human health, social development and stability. Ensuring food quality and safety is a complex process, and all stages of food processing must be considered, from cultivating, harvesting and storage to preparation and consumption. However, these processes are often labour-intensive. Nowadays, the development of machine vision can greatly assist researchers and industries in improving the efficiency of food processing. As a result, machine vision has been widely used in all aspects of food processing. At the same time, image processing is an important component of machine vision. Image processing can take advantage of machine learning and deep learning models to effectively identify the type and quality of food. Subsequently, follow-up design in the machine vision system can address tasks such as food grading, detecting locations of defective spots or foreign objects, and removing impurities. In this paper, we provide an overview on the traditional machine learning and deep learning methods, as well as the machine vision techniques that can be applied to the field of food processing. We present the current approaches and challenges, and the future trends.

Zhu Lili, Spachos Petros, Pensini Erica, Plataniotis Konstantinos N

2021

Deep learning, Food processing, Image processing, Machine learning, Machine vision

General General

Characterising the efficacy and bioavailability of bioactive peptides identified for attenuating muscle atrophy within a Vicia faba-derived functional ingredient.

In Current research in food science

Characterising key components within functional ingredients as well as assessing efficacy and bioavailability is an important step in validating nutritional interventions. Machine learning can assess large and complex data sets, such as proteomic data from plants sources, and so offers a prime opportunity to predict key bioactive components within a larger matrix. Using machine learning, we identified two potentially bioactive peptides within a Vicia faba derived hydrolysate, NPN_1, an ingredient which was previously identified for preventing muscle loss in a murine disuse model. We investigated the predicted efficacy of these peptides in vitro and observed that HLPSYSPSPQ and TIKIPAGT were capable of increasing protein synthesis and reducing TNF-α secretion, respectively. Following confirmation of efficacy, we assessed bioavailability and stability of these predicted peptides and found that as part of NPN_1, both HLPSYSPSPQ and TIKIPAGT survived upper gut digestion, were transported across the intestinal barrier and exhibited notable stability in human plasma. This work is a first step in utilising machine learning to untangle the complex nature of functional ingredients to predict active components, followed by subsequent assessment of their efficacy, bioavailability and human plasma stability in an effort to assist in the characterisation of nutritional interventions.

Corrochano Alberto R, Cal Roi, Kennedy Kathy, Wall Audrey, Murphy Niall, Trajkovic Sanja, O’Callaghan Sean, Adelfio Alessandro, Khaldi Nora

2021

Anti-inflammatory, Bioactive peptide, Intestinal absorption, Machine learning, Protein synthesis, Simulated gastrointestinal digestion

Radiology Radiology

The RSNA Pulmonary Embolism CT Dataset.

In Radiology. Artificial intelligence

Supplemental material is available for this article.

Colak Errol, Kitamura Felipe C, Hobbs Stephen B, Wu Carol C, Lungren Matthew P, Prevedello Luciano M, Kalpathy-Cramer Jayashree, Ball Robyn L, Shih George, Stein Anouk, Halabi Safwan S, Altinmakas Emre, Law Meng, Kumar Parveen, Manzalawi Karam A, Nelson Rubio Dennis Charles, Sechrist Jacob W, Germaine Pauline, Lopez Eva Castro, Amerio Tomas, Gupta Pushpender, Jain Manoj, Kay Fernando U, Lin Cheng Ting, Sen Saugata, Revels Jonathan Wesley, Brussaard Carola C, Mongan John

2021-Mar

Radiology Radiology

Automated Analysis of Alignment in Long-Leg Radiographs by Using a Fully Automated Support System Based on Artificial Intelligence.

In Radiology. Artificial intelligence

Purpose : To develop and validate a deep learning-based method for automatic quantitative analysis of lower-extremity alignment.

Materials and Methods : In this retrospective study, bilateral long-leg radiographs (LLRs) from 255 patients that were obtained between January and September of 2018 were included. For training data (n = 109), a U-Net convolutional neural network was trained to segment the femur and tibia versus manual segmentation. For validation data (n = 40), model parameters were optimized. Following identification of anatomic landmarks, anatomic and mechanical axes were identified and used to quantify alignment through the hip-knee-ankle angle (HKAA) and femoral anatomic-mechanical angle (AMA). For testing data (n = 106), algorithm-based angle measurements were compared with reference measurements by two radiologists. Angles and time for 30 random radiographs were compared by using repeated-measures analysis of variance and one-way analysis of variance, whereas correlations were quantified by using Pearson r and intraclass correlation coefficients.

Results : Bilateral LLRs of 255 patients (mean age, 26 years ± 23 [standard deviation]; range, 0-88 years; 157 male patients) were included. Mean Sørensen-Dice coefficients for segmentation were 0.97 ± 0.09 for the femur and 0.96 ± 0.11 for the tibia. Mean HKAAs and AMAs as measured by the readers and the algorithm ranged from 0.05° to 0.11° (P = .5) and from 4.82° to 5.43° (P < .001). Interreader correlation coefficients ranged from 0.918 to 0.995 (r range, P < .001), and agreement was almost perfect (intraclass correlation coefficient range, 0.87-0.99). Automatic analysis was faster than the two radiologists' manual measurements (3 vs 36 vs 35 seconds, P < .001).

Conclusion : Fully automated analysis of LLRs yielded accurate results across a wide range of clinical and pathologic indications and is fast enough to enhance and accelerate clinical workflows.Supplemental material is available for this article.© RSNA, 2020See also commentary by Andreisek in this issue.

Schock Justus, Truhn Daniel, Abrar Daniel B, Merhof Dorit, Conrad Stefan, Post Manuel, Mittelstrass Felix, Kuhl Christiane, Nebelung Sven

2021-Mar

Radiology Radiology

Automated Segmentation of Abdominal Skeletal Muscle on Pediatric CT Scans Using Deep Learning.

In Radiology. Artificial intelligence

Purpose : To automate skeletal muscle segmentation in a pediatric population using convolutional neural networks that identify and segment the L3 level at CT.

Materials and Methods : In this retrospective study, two sets of U-Net-based models were developed to identify the L3 level in the sagittal plane and segment the skeletal muscle from the corresponding axial image. For model development, 370 patients (sampled uniformly across age group from 0 to 18 years and including both sexes) were selected between January 2009 and January 2019, and ground truth L3 location and skeletal muscle segmentation were manually defined. Twenty percent (74 of 370) of the examinations were reserved for testing the L3 locator and muscle segmentation, while the remaining were used for training. For the L3 locator models, maximum intensity projections (MIPs) from a fixed number of central sections of sagittal reformats (either 12 or 18 sections) were used as input with or without transfer learning using an L3 localizer trained on an external dataset (four models total). For the skeletal muscle segmentation models, two loss functions (weighted Dice similarity coefficient [DSC] and binary cross-entropy) were used on models trained with or without data augmentation (four models total). Outputs from each model were compared with ground truth, and the mean relative error and DSC from each of the models were compared with one another.

Results : L3 section detection trained with an 18-section MIP model with transfer learning had a mean error of 3.23 mm ± 2.61 standard deviation, which was within the reconstructed image thickness (3 or 5 mm). Skeletal muscle segmentation trained with the weighted DSC loss model without data augmentation had a mean DSC of 0.93 ± 0.03 and mean relative error of 0.04 ± 0.04.

Conclusion : Convolutional neural network models accurately identified the L3 level and segmented the skeletal muscle on pediatric CT scans.Supplemental material is available for this article.See also the commentary by Cadrin-Chênevert in this issue.© RSNA, 2021.

Castiglione James, Somasundaram Elanchezhian, Gilligan Leah A, Trout Andrew T, Brady Samuel

2021-Mar

General General

Medical image-based detection of COVID-19 using Deep Convolution Neural Networks.

In Multimedia systems

The demand for automatic detection of Novel Coronavirus or COVID-19 is increasing across the globe. The exponential rise in cases burdens healthcare facilities, and a vast amount of multimedia healthcare data is being explored to find a solution. This study presents a practical solution to detect COVID-19 from chest X-rays while distinguishing those from normal and impacted by Viral Pneumonia via Deep Convolution Neural Networks (CNN). In this study, three pre-trained CNN models (EfficientNetB0, VGG16, and InceptionV3) are evaluated through transfer learning. The rationale for selecting these specific models is their balance of accuracy and efficiency with fewer parameters suitable for mobile applications. The dataset used for the study is publicly available and compiled from different sources. This study uses deep learning techniques and performance metrics (accuracy, recall, specificity, precision, and F1 scores). The results show that the proposed approach produced a high-quality model, with an overall accuracy of 92.93%, COVID-19, a sensitivity of 94.79%. The work indicates a definite possibility to implement computer vision design to enable effective detection and screening measures.

Gaur Loveleen, Bhatia Ujwal, Jhanjhi N Z, Muhammad Ghulam, Masud Mehedi

2021-Apr-28

COVID-19, Chest X-rays, Computer vision, Deep CNN, Deep learning, Transfer learning

General General

Classification of COVID-19 Chest CT Images Based on Ensemble Deep Learning.

In Journal of healthcare engineering

Novel coronavirus pneumonia (NCP) has become a global pandemic disease, and computed tomography-based (CT) image analysis and recognition are one of the important tools for clinical diagnosis. In order to assist medical personnel to achieve an efficient and fast diagnosis of patients with new coronavirus pneumonia, this paper proposes an assisted diagnosis algorithm based on ensemble deep learning. The method combines the Stacked Generalization ensemble learning with the VGG16 deep learning to form a cascade classifier, and the information constituting the cascade classifier comes from multiple subsets of the training set, each of which is used to collect deviant information about the generalization behavior of the data set, such that this deviant information fills the cascade classifier. The algorithm was experimentally validated for classifying patients with novel coronavirus pneumonia, patients with common pneumonia (CP), and normal controls, and the algorithm achieved a prediction accuracy of 93.57%, sensitivity of 94.21%, specificity of 93.93%, precision of 89.40%, and F1-score of 91.74% for the three categories. The results show that the method proposed in this paper has good classification performance and can significantly improve the performance of deep neural networks for multicategory prediction tasks.

Li Xiaoshuo, Tan Wenjun, Liu Pan, Zhou Qinghua, Yang Jinzhu

2021

Radiology Radiology

Active Reprioritization of the Reading Worklist Using Artificial Intelligence Has a Beneficial Effect on the Turnaround Time for Interpretation of Head CT with Intracranial Hemorrhage.

In Radiology. Artificial intelligence

Purpose : To determine how to optimize the delivery of machine learning techniques in a clinical setting to detect intracranial hemorrhage (ICH) on non-contrast-enhanced CT images to radiologists to improve workflow.

Materials and Methods : In this study, a commercially available machine learning algorithm that flags abnormal noncontrast CT examinations for ICH was implemented in a busy academic neuroradiology practice between September 2017 and March 2019. The algorithm was introduced in three phases: (a) as a "pop-up" widget on ancillary monitors, (b) as a marked examination in reading worklists, and (c) as a marked examination for reprioritization based on the presence of the flag. A statistical approach, which was based on a queuing theory, was implemented to assess the impact of each intervention on queue-adjusted wait and turnaround time compared with historical controls.

Results : Notification with a widget or flagging the examination had no effect on queue-adjusted image wait (P > .99) or turnaround time (P = .6). However, a reduction in queue-adjusted wait time was observed between negative (15.45 minutes; 95% CI: 15.07, 15.38) and positive (12.02 minutes; 95% CI: 11.06, 12.97; P < .0001) artificial intelligence-detected ICH examinations with reprioritization. Reduced wait time was present for all order classes but was greatest for examinations ordered as routine for both inpatients and outpatients because of their low priority.

Conclusion : The approach used to present flags from artificial intelligence and machine learning algorithms to the radiologist can reduce image wait time and turnaround times.© RSNA, 2021See also the commentary by O'Connor and Bhalla in this issue.

O’Neill Thomas J, Xi Yin, Stehel Edward, Browning Travis, Ng Yee Seng, Baker Chris, Peshock Ronald M

2021-Mar

Radiology Radiology

Deep Learning to Quantify Pulmonary Edema in Chest Radiographs.

In Radiology. Artificial intelligence

Purpose : To develop a machine learning model to classify the severity grades of pulmonary edema on chest radiographs.

Materials and Methods : In this retrospective study, 369 071 chest radiographs and associated radiology reports from 64 581 patients (mean age, 51.71 years; 54.51% women) from the MIMIC-CXR chest radiograph dataset were included. This dataset was split into patients with and without congestive heart failure (CHF). Pulmonary edema severity labels from the associated radiology reports were extracted from patients with CHF as four different ordinal levels: 0, no edema; 1, vascular congestion; 2, interstitial edema; and 3, alveolar edema. Deep learning models were developed using two approaches: a semisupervised model using a variational autoencoder and a pretrained supervised learning model using a dense neural network. Receiver operating characteristic curve analysis was performed on both models.

Results : The area under the receiver operating characteristic curve (AUC) for differentiating alveolar edema from no edema was 0.99 for the semisupervised model and 0.87 for the pretrained models. Performance of the algorithm was inversely related to the difficulty in categorizing milder states of pulmonary edema (shown as AUCs for semisupervised model and pretrained model, respectively): 2 versus 0, 0.88 and 0.81; 1 versus 0, 0.79 and 0.66; 3 versus 1, 0.93 and 0.82; 2 versus 1, 0.69 and 0.73; and 3 versus 2, 0.88 and 0.63.

Conclusion : Deep learning models were trained on a large chest radiograph dataset and could grade the severity of pulmonary edema on chest radiographs with high performance.Supplemental material is available for this article.See also the commentary by Auffermann in this issue.© RSNA, 2021.

Horng Steven, Liao Ruizhi, Wang Xin, Dalal Sandeep, Golland Polina, Berkowitz Seth J

2021-Mar

Radiology Radiology

Artificial Intelligence for Classification of Soft-Tissue Masses at US.

In Radiology. Artificial intelligence

Purpose : To train convolutional neural network (CNN) models to classify benign and malignant soft-tissue masses at US and to differentiate three commonly observed benign masses.

Materials and Methods : In this retrospective study, US images obtained between May 2010 and June 2019 from 419 patients (mean age, 52 years ± 18 [standard deviation]; 250 women) with histologic diagnosis confirmed at biopsy or surgical excision (n = 227) or masses that demonstrated imaging characteristics of lipoma, benign peripheral nerve sheath tumor, and vascular malformation (n = 192) were included. Images in patients with a histologic diagnosis (n = 227) were used to train and evaluate a CNN model to distinguish malignant and benign lesions. Twenty percent of cases were withheld as a test dataset, and the remaining cases were used to train the model with a 75%-25% training-validation split and fourfold cross-validation. Performance of the model was compared with retrospective interpretation of the same dataset by two experienced musculoskeletal radiologists, blinded to clinical history. A second group of US images from 275 of the 419 patients containing the three common benign masses was used to train and evaluate a separate model to differentiate between the masses. The models were trained on the Keras machine learning platform (version 2.3.1), with a modified pretrained VGG16 network. Performance metrics of the model and of the radiologists were compared by using the McNemar test, and 95% CIs for performance metrics were estimated by using the Clopper-Pearson method (accuracy, recall, specificity, and precision) and the DeLong method (area under the receiver operating characteristic curve).

Results : The model trained to classify malignant and benign masses demonstrated an accuracy of 79% (95% CI: 68, 88) on the test data, with an area under the receiver operating characteristic curve of 0.91 (95% CI: 0.84, 0.98), matching the performance of two expert readers. Performance of the model distinguishing three benign masses was lower, with an accuracy of 71% (95% CI: 61, 80) on the test data.

Conclusion : The trained CNN was capable of differentiating between benign and malignant soft-tissue masses depicted on US images, with performance matching that of two experienced musculoskeletal radiologists.© RSNA, 2020.

Wang Benjamin, Perronne Laetitia, Burke Christopher, Adler Ronald S

2021-Jan

Radiology Radiology

Magician's Corner: 8: How to Connect an Artificial Intelligence Tool to PACS.

In Radiology. Artificial intelligence

Authors show how to use Digital Imaging and Communications in Medicine (DICOM) Query and Retrieve functions to pull a study from a cloud or public picture archiving and communication system (PACS), run an artificial intelligence (AI) algorithm on those images, and store the results back to another (cloud) PACS.

Erickson Bradley J, Kitamura Felipe

2021-Jan

Radiology Radiology

Endotracheal Tube Position Assessment on Chest Radiographs Using Deep Learning.

In Radiology. Artificial intelligence

Purpose : To determine the efficacy of deep learning in assessing endotracheal tube (ETT) position on radiographs.

Materials and Methods : In this retrospective study, 22 960 de-identified frontal chest radiographs from 11 153 patients (average age, 60.2 years ± 19.9 [standard deviation], 55.6% men) between 2010 and 2018 containing an ETT were placed into 12 categories, including bronchial insertion and distance from the carina at 1.0-cm intervals (0.0-0.9 cm, 1.0-1.9 cm, etc), and greater than 10 cm. Images were split into training (80%, 18 368 images), validation (10%, 2296 images), and internal test (10%, 2296 images), derived from the same institution as the training data. One hundred external test radiographs were also obtained from a different hospital. The Inception V3 deep neural network was used to predict ETT-carina distance. ETT-carina distances and intraclass correlation coefficients (ICCs) for the radiologists and artificial intelligence (AI) system were calculated on a subset of 100 random internal and 100 external test images. Sensitivity and specificity were calculated for low and high ETT position thresholds.

Results : On the internal and external test images, respectively, the ICCs of AI and radiologists were 0.84 (95% CI: 0.78, 0.92) and 0.89 (95% CI: 0.77, 0.94); the ICCs of the radiologists were 0.93 (95% CI: 0.90, 0.95) and 0.84 (95% CI: 0.71, 0.90). The AI model was 93.9% sensitive (95% CI: 90.0, 96.7) and 97.7% specific (95% CI: 96.9, 98.3) for detecting ETT-carina distance less than 1 cm.

Conclusion : Deep learning predicted ETT-carina distance within 1 cm in most cases and showed excellent interrater agreement compared with radiologists. The model was sensitive and specific in detecting low ETT positions.© RSNA, 2020.

Lakhani Paras, Flanders Adam, Gorniak Richard

2021-Jan

Radiology Radiology

Deep Learning-based Automated Segmentation of Left Ventricular Trabeculations and Myocardium on Cardiac MR Images: A Feasibility Study.

In Radiology. Artificial intelligence

Purpose : To develop and evaluate a complete deep learning pipeline that allows fully automated end-diastolic left ventricle (LV) cardiac MRI segmentation, including trabeculations and automatic quality control of the predicted segmentation.

Materials and Methods : This multicenter retrospective study includes training, validation, and testing datasets of 272, 27, and 150 cardiac MR images, respectively, collected between 2012 and 2018. The reference standard was the manual segmentation of four LV anatomic structures performed on end-diastolic short-axis cine cardiac MRI: LV trabeculations, LV myocardium, LV papillary muscles, and the LV blood cavity. The automatic pipeline was composed of five steps with a DenseNet architecture. Intraobserver agreement, interobserver agreement, and interaction time were recorded. The analysis includes the correlation between the manual and automated segmentation, a reproducibility comparison, and Bland-Altman plots.

Results : The automated method achieved mean Dice coefficients of 0.96 ± 0.01 (standard deviation) for LV blood cavity, 0.89 ± 0.03 for LV myocardium, and 0.62 ± 0.08 for LV trabeculation (mean absolute error, 3.63 g ± 3.4). Automatic quantification of LV end-diastolic volume, LV myocardium mass, LV trabeculation, and trabeculation mass-to-total myocardial mass (TMM) ratio showed a significant correlation with the manual measures (r = 0.99, 0.99, 0.90, and 0.83, respectively; all P < .01). On a subset of 48 patients, the mean Dice value for LV trabeculation was 0.63 ± 0.10 or higher compared with the human interobserver (0.44 ± 0.09; P < .01) and intraobserver measures (0.58 ± 0.09; P < .01). Automatic quantification of the trabeculation mass-to-TMM ratio had a higher correlation (0.92) compared with the intra- and interobserver measures (0.74 and 0.39, respectively; both P < .01).

Conclusion : Automated deep learning framework can achieve reproducible and quality-controlled segmentation of cardiac trabeculations, outperforming inter- and intraobserver analyses.Supplemental material is available for this article.© RSNA, 2020.

Bartoli Axel, Fournel Joris, Bentatou Zakarya, Habib Gilbert, Lalande Alain, Bernard Monique, Boussel Loïc, Pontana François, Dacher Jean-Nicolas, Ghattas Badih, Jacquier Alexis

2021-Jan

Radiology Radiology

A Multisite Study of a Breast Density Deep Learning Model for Full-Field Digital Mammography and Synthetic Mammography.

In Radiology. Artificial intelligence

Purpose : To develop a Breast Imaging Reporting and Data System (BI-RADS) breast density deep learning (DL) model in a multisite setting for synthetic two-dimensional mammographic (SM) images derived from digital breast tomosynthesis examinations by using full-field digital mammographic (FFDM) images and limited SM data.

Materials and Methods : A DL model was trained to predict BI-RADS breast density by using FFDM images acquired from 2008 to 2017 (site 1: 57 492 patients, 187 627 examinations, 750 752 images) for this retrospective study. The FFDM model was evaluated by using SM datasets from two institutions (site 1: 3842 patients, 3866 examinations, 14 472 images, acquired from 2016 to 2017; site 2: 7557 patients, 16 283 examinations, 63 973 images, 2015 to 2019). Each of the three datasets were then split into training, validation, and test. Adaptation methods were investigated to improve performance on the SM datasets, and the effect of dataset size on each adaptation method was considered. Statistical significance was assessed by using CIs, which were estimated by bootstrapping.

Results : Without adaptation, the model demonstrated substantial agreement with the original reporting radiologists for all three datasets (site 1 FFDM: linearly weighted Cohen κ [κw] = 0.75 [95% CI: 0.74, 0.76]; site 1 SM: κw = 0.71 [95% CI: 0.64, 0.78]; site 2 SM: κw = 0.72 [95% CI: 0.70, 0.75]). With adaptation, performance improved for site 2 (site 1: κw = 0.72 [95% CI: 0.66, 0.79], 0.71 vs 0.72, P = .80; site 2: κw = 0.79 [95% CI: 0.76, 0.81], 0.72 vs 0.79, P < .001) by using only 500 SM images from that site.

Conclusion : A BI-RADS breast density DL model demonstrated strong performance on FFDM and SM images from two institutions without training on SM images and improved by using few SM images.Supplemental material is available for this article.Published under a CC BY 4.0 license.

Matthews Thomas P, Singh Sadanand, Mombourquette Brent, Su Jason, Shah Meet P, Pedemonte Stefano, Long Aaron, Maffit David, Gurney Jenny, Hoil Rodrigo Morales, Ghare Nikita, Smith Douglas, Moore Stephen M, Marks Susan C, Wahl Richard L

2021-Jan

Radiology Radiology

Preparing Radiologists to Lead in the Era of Artificial Intelligence: Designing and Implementing a Focused Data Science Pathway for Senior Radiology Residents.

In Radiology. Artificial intelligence

Artificial intelligence and machine learning (AI-ML) have taken center stage in medical imaging. To develop as leaders in AI-ML, radiology residents may seek a formative data science experience. The authors piloted an elective Data Science Pathway (DSP) for 4th-year residents at the authors' institution in collaboration with the MGH & BWH Center for Clinical Data Science (CCDS). The goal of the DSP was to provide an introduction to AI-ML through a flexible schedule of educational, experiential, and research activities. The study describes the initial experience with the DSP tailored to the AI-ML interests of three senior radiology residents. The authors also discuss logistics and curricular design with common core elements and shared mentorship. Residents were provided dedicated, full-time immersion into the CCDS work environment. In the initial DSP pilot, residents were successfully integrated into AI-ML projects at CCDS. Residents were exposed to all aspects of AI-ML application development, including data curation, model design, quality control, and clinical testing. Core concepts in AI-ML were taught through didactic sessions and daily collaboration with data scientists and other staff. Work during the pilot period led to 12 accepted abstracts for presentation at national meetings. The DSP is a feasible, well-rounded introductory experience in AI-ML for senior radiology residents. Residents contributed to model and tool development at multiple stages and were academically productive. Feedback from the pilot resulted in establishment of a formal AI-ML curriculum for future residents. The described logistical, planning, and curricular considerations provide a framework for DSP implementation at other institutions. Supplemental material is available for this article. © RSNA, 2020.

Wiggins Walter F, Caton M Travis, Magudia Kirti, Glomski Sha-Har A, George Elizabeth, Rosenthal Michael H, Gaviola Glenn C, Andriole Katherine P

2020-Nov

Radiology Radiology

Fully Automated and Standardized Segmentation of Adipose Tissue Compartments via Deep Learning in 3D Whole-Body MRI of Epidemiologic Cohort Studies.

In Radiology. Artificial intelligence

Purpose : To enable fast and reliable assessment of subcutaneous and visceral adipose tissue compartments derived from whole-body MRI.

Materials and Methods : Quantification and localization of different adipose tissue compartments derived from whole-body MR images is of high interest in research concerning metabolic conditions. For correct identification and phenotyping of individuals at increased risk for metabolic diseases, a reliable automated segmentation of adipose tissue into subcutaneous and visceral adipose tissue is required. In this work, a three-dimensional (3D) densely connected convolutional neural network (DCNet) is proposed to provide robust and objective segmentation. In this retrospective study, 1000 cases (average age, 66 years ± 13 [standard deviation]; 523 women) from the Tuebingen Family Study database and the German Center for Diabetes research database and 300 cases (average age, 53 years ± 11; 152 women) from the German National Cohort (NAKO) database were collected for model training, validation, and testing, with transfer learning between the cohorts. These datasets included variable imaging sequences, imaging contrasts, receiver coil arrangements, scanners, and imaging field strengths. The proposed DCNet was compared to a similar 3D U-Net segmentation in terms of sensitivity, specificity, precision, accuracy, and Dice overlap.

Results : Fast (range, 5-7 seconds) and reliable adipose tissue segmentation can be performed with high Dice overlap (0.94), sensitivity (96.6%), specificity (95.1%), precision (92.1%), and accuracy (98.4%) from 3D whole-body MRI datasets (field of view coverage, 450 × 450 × 2000 mm). Segmentation masks and adipose tissue profiles are automatically reported back to the referring physician.

Conclusion : Automated adipose tissue segmentation is feasible in 3D whole-body MRI datasets and is generalizable to different epidemiologic cohort studies with the proposed DCNet.Supplemental material is available for this article.© RSNA, 2020.

Küstner Thomas, Hepp Tobias, Fischer Marc, Schwartz Martin, Fritsche Andreas, Häring Hans-Ulrich, Nikolaou Konstantin, Bamberg Fabian, Yang Bin, Schick Fritz, Gatidis Sergios, Machann Jürgen

2020-Nov

Radiology Radiology

The State of Radiology AI: Considerations for Purchase Decisions and Current Market Offerings.

In Radiology. Artificial intelligence

Purpose : To provide an overview of important factors to consider when purchasing radiology artificial intelligence (AI) software and current software offerings by type, subspecialty, and modality.

Materials and Methods : Important factors for consideration when purchasing AI software, including key decision makers, data ownership and privacy, cost structures, performance indicators, and potential return on investment are described. For the market overview, a list of radiology AI companies was aggregated from the Radiological Society of North America and the Society for Imaging Informatics in Medicine conferences (November 2016-June 2019), then narrowed to companies using deep learning for imaging analysis and diagnosis. Software created for image enhancement, reporting, or workflow management was excluded. Software was categorized by task (repetitive, quantitative, explorative, and diagnostic), modality, and subspecialty.

Results : A total of 119 software offerings from 55 companies were identified. There were 46 algorithms that currently have Food and Drug Administration and/or Conformité Européenne approval (as of November 2019). Of the 119 offerings, distribution of software targets was 34 of 70 (49%), 21 of 70 (30%), 14 of 70 (20%), and one of 70 (1%) for diagnostic, quantitative, repetitive, and explorative tasks, respectively. A plurality of companies are focused on nodule detection at chest CT and two-dimensional mammography. There is very little activity in certain subspecialties, including pediatrics and nuclear medicine. A comprehensive table is available on the website hitilab.org/pages/ai-companies.

Conclusion : The radiology AI marketplace is rapidly maturing, with an increase in product offerings. Radiologists and practice administrators should educate themselves on current product offerings and important factors to consider before purchase and implementation.© RSNA, 2020See also the invited commentary by Sala and Ursprung in this issue.

Tadavarthi Yasasvi, Vey Brianna, Krupinski Elizabeth, Prater Adam, Gichoya Judy, Safdar Nabile, Trivedi Hari

2020-Nov

General General

Bifurcation and chaos analysis for a discrete ecological developmental systems.

In Nonlinear dynamics

This work concentrates on the dynamic analysis including bifurcation and chaos of a discrete ecological developmental systems. Specifically, it is a prey-predator-scavenger (PPS) system, which is derived by Euler discretization method. By choosing the step size h as a bifurcation parameter, we determine the set consists of all system's parameters, in which the system can undergo flip bifurcation (FB) and Neimark-Sacker bifurcation (NSB). The theoretical results are verified by some numerical simulations. It is shown that the discrete systems exhibit more interesting behaviors, including the chaotic sets, quasi-periodic orbits, and the cascade of period-doubling bifurcation in orbits of periods 2, 4, 8, 16. Finally, corresponding to the two bifurcation behaviors discussed, the maximum Lyapunov exponent is numerically calculated, which further verifies the rich dynamic characteristics of the discrete system.

Jiang Xiao-Wei, Chen Chaoyang, Zhang Xian-He, Chi Ming, Yan Huaicheng

2021-Apr-26

Chaos, Ecological developmental systems, Flip bifurcation, Neimark–Sacker bifurcation, Stability

Public Health Public Health

Leveraging Informatics and Technology to Support Public Health Response: Framework and Illustrations using COVID-19.

In Online journal of public health informatics

Objective : To develop a conceptual model and novel, comprehensive framework that encompass the myriad ways informatics and technology can support public health response to a pandemic.

Method : The conceptual model and framework categorize informatics solutions that could be used by stakeholders (e.g., government, academic institutions, healthcare providers and payers, life science companies, employers, citizens) to address public health challenges across the prepare, respond, and recover phases of a pandemic, building on existing models for public health operations and response.

Results : Mapping existing solutions, technology assets, and ideas to the framework helped identify public health informatics solution requirements and gaps in responding to COVID-19 in areas such as applied science, epidemiology, communications, and business continuity. Two examples of technologies used in COVID-19 illustrate novel applications of informatics encompassed by the framework. First, we examine a hub from The Weather Channel, which provides COVID-19 data via interactive maps, trend graphs, and details on case data to individuals and businesses. Second, we examine IBM Watson Assistant for Citizens, an AI-powered virtual agent implemented by healthcare providers and payers, government agencies, and employers to provide information about COVID-19 via digital and telephone-based interaction.

Discussion : Early results from these novel informatics solutions have been positive, showing high levels of engagement and added value across stakeholders.

Conclusion : The framework supports development, application, and evaluation of informatics approaches and technologies in support of public health preparedness, response, and recovery during a pandemic. Effective solutions are critical to success in recovery from COVID-19 and future pandemics.

Snowdon Jane L, Kassler William, Karunakaram Hema, Dixon Brian E, Rhee Kyu

2021

artificial intelligence, clinical informatics, coronavirus, information technology, pandemics, public health informatics

Radiology Radiology

Improving Breast Cancer Detection Accuracy of Mammography with the Concurrent Use of an Artificial Intelligence Tool.

In Radiology. Artificial intelligence

Purpose : To evaluate the benefits of an artificial intelligence (AI)-based tool for two-dimensional mammography in the breast cancer detection process.

Materials and Methods : In this multireader, multicase retrospective study, 14 radiologists assessed a dataset of 240 digital mammography images, acquired between 2013 and 2016, using a counterbalance design in which half of the dataset was read without AI and the other half with the help of AI during a first session and vice versa during a second session, which was separated from the first by a washout period. Area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and reading time were assessed as endpoints.

Results : The average AUC across readers was 0.769 (95% CI: 0.724, 0.814) without AI and 0.797 (95% CI: 0.754, 0.840) with AI. The average difference in AUC was 0.028 (95% CI: 0.002, 0.055, P = .035). Average sensitivity was increased by 0.033 when using AI support (P = .021). Reading time changed dependently to the AI-tool score. For low likelihood of malignancy (< 2.5%), the time was about the same in the first reading session and slightly decreased in the second reading session. For higher likelihood of malignancy, the reading time was on average increased with the use of AI.

Conclusion : This clinical investigation demonstrated that the concurrent use of this AI tool improved the diagnostic performance of radiologists in the detection of breast cancer without prolonging their workflow.Supplemental material is available for this article.© RSNA, 2020.

Pacilè Serena, Lopez January, Chone Pauline, Bertinotti Thomas, Grouin Jean Marie, Fillard Pierre

2020-Nov

Radiology Radiology

Evaluation of Automated Public De-Identification Tools on a Corpus of Radiology Reports.

In Radiology. Artificial intelligence

Purpose : To evaluate publicly available de-identification tools on a large corpus of narrative-text radiology reports.

Materials and Methods : In this retrospective study, 21 categories of protected health information (PHI) in 2503 radiology reports were annotated from a large multihospital academic health system, collected between January 1, 2012 and January 8, 2019. A subset consisting of 1023 reports served as a test set; the remainder were used as domain-specific training data. The types and frequencies of PHI present within the reports were tallied. Five public de-identification tools were evaluated: MITRE Identification Scrubber Toolkit, U.S. National Library of Medicine‒Scrubber, Massachusetts Institute of Technology de-identification software, Emory Health Information DE-identification (HIDE) software, and Neuro named-entity recognition (NeuroNER). The tools were compared using metrics including recall, precision, and F1 score (the harmonic mean of recall and precision) for each category of PHI.

Results : The annotators identified 3528 spans of PHI text within the 2503 reports. Cohen κ for interrater agreement was 0.938. Dates accounted for the majority of PHI found in the dataset of radiology reports (n = 2755 [78%]). The two best-performing tools both used machine learning methods-NeuroNER (precision, 94.5%; recall, 92.6%; microaveraged F1 score [F1], 93.6%) and Emory HIDE (precision, 96.6%; recall, 88.2%; F1, 92.2%)-but none exceeded 50% F1 on the important patient names category.

Conclusion : PHI appeared infrequently within the corpus of reports studied, which created difficulties for training machine learning systems. Out-of-the-box de-identification tools achieved limited performance on the corpus of radiology reports, suggesting the need for further advancements in public datasets and trained models.Supplemental material is available for this article.See also the commentary by Tenenholtz and Wood in this issue.© RSNA, 2020.

Steinkamp Jackson M, Pomeranz Taylor, Adleberg Jason, Kahn Charles E, Cook Tessa S

2020-Nov

Radiology Radiology

Automatic Detection of Inadequate Pediatric Lateral Neck Radiographs of the Airway and Soft Tissues using Deep Learning.

In Radiology. Artificial intelligence

Purpose : To develop and validate a deep learning (DL) algorithm to identify poor-quality lateral airway radiographs.

Materials and Methods : A total of 1200 lateral airway radiographs obtained in emergency department patients between January 1, 2000, and July 1, 2019, were retrospectively queried from the picture archiving and communication system. Two radiologists classified each radiograph as adequate or inadequate. Disagreements were adjudicated by a third radiologist. The radiographs were used to train and test the DL classifiers. Three technologists and three different radiologists classified the images in the test dataset, and their performance was compared with that of the DL classifiers.

Results : The training set had 961 radiographs and the test set had 239. The best DL classifier (ResNet-50) achieved sensitivity, specificity, and area under the receiver operating characteristic curve of 0.90 (95% confidence interval [CI]: 0.86, 0.94), 0.82 (95% CI: 0.76, 0.90), and 0.86 (95% CI: 0.81, 0.91), respectively. Interrater agreement for technologists was fair (Fleiss κ, 0.36 [95% CI: 0.29, 0.43]), while that for radiologists was moderate (Fleiss κ, 0.59 [95% CI: 0.52, 0.66]). Cohen κ value comparing the consensus rating of ResNet-50 iterations from fivefold cross-validation, consensus technologists' rating, and consensus radiologists' rating to the ground truth were 0.76 (95% CI: 0.63, 0.89), 0.49 (95% CI: 0.37, 0.61), and 0.66 (95% CI: 0.54, 0.78), respectively.

Conclusion : The development and validation of DL classifiers to distinguish between adequate and inadequate lateral airway radiographs is reported. The classifiers performed significantly better than a group of technologists and as well as the radiologists.© RSNA, 2020.

Somasundaram Elanchezhian, Dillman Jonathan R, Crotty Eric J, Trout Andrew T, Towbin Alexander J, Anton Christopher G, Logan Angeline, Wieland Catherine A, Felekey Samantha, Coley Brian D, Brady Samuel L

2020-Sep

Radiology Radiology

Improved Segmentation and Detection Sensitivity of Diffusion-weighted Stroke Lesions with Synthetically Enhanced Deep Learning.

In Radiology. Artificial intelligence

Purpose : To compare the segmentation and detection performance of a deep learning model trained on a database of human-labeled clinical stroke lesions on diffusion-weighted (DW) images to a model trained on the same database enhanced with synthetic stroke lesions.

Materials and Methods : In this institutional review board-approved study, a stroke database of 962 cases (mean patient age ± standard deviation, 65 years ± 17; 255 male patients; 449 scans with DW positive stroke lesions) and a normal database of 2027 patients (mean age, 38 years ± 24; 1088 female patients) were used. Brain volumes with synthetic stroke lesions on DW images were produced by warping the relative signal increase of real strokes to normal brain volumes. A generic three-dimensional (3D) U-Net was trained on four different databases to generate four different models: (a) 375 neuroradiologist-labeled clinical DW positive stroke cases (CDB); (b) 2000 synthetic cases (S2DB); (c) CDB plus 2000 synthetic cases (CS2DB); and (d) CDB plus 40 000 synthetic cases (CS40DB). The models were tested on 20% (n = 192) of the cases of the stroke database, which were excluded from the training set. Segmentation accuracy was characterized using Dice score and lesion volume of the stroke segmentation, and statistical significance was tested using a paired two-tailed Student t test. Detection sensitivity and specificity were compared with labeling done by three neuroradiologists.

Results : The performance of the 3D U-Net model trained on the CS40DB (mean Dice score, 0.72) was better than models trained on the CS2DB (Dice score, 0.70; P < .001) or the CDB (Dice score, 0.65; P < .001). The deep learning model (CS40DB) was also more sensitive (91% [95% confidence interval {CI}: 89%, 93%]) than each of the three human readers (human reader 3, 84% [95% CI: 81%, 87%]; human reader 1, 78% [95% CI: 75%, 81%]; human reader 2, 79% [95% CI: 76%, 82%]), but was less specific (75% [95% CI: 72%, 78%]) than each of the three human readers (human reader 3, 96% [95% CI: 94%, 98%]; human reader 1, 92% [95% CI: 90%, 94%]; human reader 2, 89% [95% CI: 86%, 91%]).

Conclusion : Deep learning training for segmentation and detection of stroke lesions on DW images was significantly improved by enhancing the training set with synthetic lesions.Supplemental material is available for this article.© RSNA, 2020.

Federau Christian, Christensen Soren, Scherrer Nino, Ospel Johanna M, Schulze-Zachau Victor, Schmidt Noemi, Breit Hanns-Christian, Maclaren Julian, Lansberg Maarten, Kozerke Sebastian

2020-Sep

Radiology Radiology

Fully Automated Segmentation of Head CT Neuroanatomy Using Deep Learning.

In Radiology. Artificial intelligence

Purpose : To develop a deep learning model that segments intracranial structures on head CT scans.

Materials and Methods : In this retrospective study, a primary dataset containing 62 normal noncontrast head CT scans from 62 patients (mean age, 73 years; age range, 27-95 years) acquired between August and December 2018 was used for model development. Eleven intracranial structures were manually annotated on the axial oblique series. The dataset was split into 40 scans for training, 10 for validation, and 12 for testing. After initial training, eight model configurations were evaluated on the validation dataset and the highest performing model was evaluated on the test dataset. Interobserver variability was reported using multirater consensus labels obtained from the test dataset. To ensure that the model learned generalizable features, it was further evaluated on two secondary datasets containing 12 volumes with idiopathic normal pressure hydrocephalus (iNPH) and 30 normal volumes from a publicly available source. Statistical significance was determined using categorical linear regression with P < .05.

Results : Overall Dice coefficient on the primary test dataset was 0.84 ± 0.05 (standard deviation). Performance ranged from 0.96 ± 0.01 (brainstem and cerebrum) to 0.74 ± 0.06 (internal capsule). Dice coefficients were comparable to expert annotations and exceeded those of existing segmentation methods. The model remained robust on external CT scans and scans demonstrating ventricular enlargement. The use of within-network normalization and class weighting facilitated learning of underrepresented classes.

Conclusion : Automated segmentation of CT neuroanatomy is feasible with a high degree of accuracy. The model generalized to external CT scans as well as scans demonstrating iNPH.Supplemental material is available for this article.© RSNA, 2020.

Cai Jason C, Akkus Zeynettin, Philbrick Kenneth A, Boonrod Arunnit, Hoodeshenas Safa, Weston Alexander D, Rouzrokh Pouria, Conte Gian Marco, Zeinoddini Atefeh, Vogelsang David C, Huang Qiao, Erickson Bradley J

2020-Sep

Radiology Radiology

Subspecialty-Level Deep Gray Matter Differential Diagnoses with Deep Learning and Bayesian Networks on Clinical Brain MRI: A Pilot Study.

In Radiology. Artificial intelligence

Purpose : To develop and validate a system that could perform automated diagnosis of common and rare neurologic diseases involving deep gray matter on clinical brain MRI studies.

Materials and Methods : In this retrospective study, multimodal brain MRI scans from 212 patients (mean age, 55 years ± 17 [standard deviation]; 113 women) with 35 neurologic diseases and normal brain MRI scans obtained between January 2008 and January 2018 were included (110 patients in the training set, 102 patients in the test set). MRI scans from 178 patients (mean age, 48 years ± 17; 106 women) were used to supplement training of the neural networks. Three-dimensional convolutional neural networks and atlas-based image processing were used for extraction of 11 imaging features. Expert-derived Bayesian networks incorporating domain knowledge were used for differential diagnosis generation. The performance of the artificial intelligence (AI) system was assessed by comparing diagnostic accuracy with that of radiologists of varying levels of specialization by using the generalized estimating equation with robust variance estimator for the top three differential diagnoses (T3DDx) and the correct top diagnosis (TDx), as well as with receiver operating characteristic analyses.

Results : In the held-out test set, the imaging pipeline detected 11 key features on brain MRI scans with 89% accuracy (sensitivity, 81%; specificity, 95%) relative to academic neuroradiologists. The Bayesian network, integrating imaging features with clinical information, had an accuracy of 85% for T3DDx and 64% for TDx, which was better than that of radiology residents (n = 4; 56% for T3DDx, 36% for TDx; P < .001 for both) and general radiologists (n = 2; 53% for T3DDx, 31% for TDx; P < .001 for both). The accuracy of the Bayesian network was better than that of neuroradiology fellows (n = 2) for T3DDx (72%; P = .003) but not for TDx (59%; P = .19) and was not different from that of academic neuroradiologists (n = 2; 84% T3DDx, 65% TDx; P > .09 for both).

Conclusion : A hybrid AI system was developed that simultaneously provides a quantitative assessment of disease burden, explainable intermediate imaging features, and a probabilistic differential diagnosis that performed at the level of academic neuroradiologists. This type of approach has the potential to improve clinical decision making for common and rare diseases.Supplemental material is available for this article.© RSNA, 2020.

Rudie Jeffrey D, Rauschecker Andreas M, Xie Long, Wang Jiancong, Duong Michael Tran, Botzolakis Emmanuel J, Kovalovich Asha, Egan John M, Cook Tessa, Bryan R Nick, Nasrallah Ilya M, Mohan Suyash, Gee James C

2020-Sep

Radiology Radiology

Robust Deep Learning-based Segmentation of Glioblastoma on Routine Clinical MRI Scans Using Sparsified Training.

In Radiology. Artificial intelligence

Purpose : To improve the robustness of deep learning-based glioblastoma segmentation in a clinical setting with sparsified datasets.

Materials and Methods : In this retrospective study, preoperative T1-weighted, T2-weighted, T2-weighted fluid-attenuated inversion recovery, and postcontrast T1-weighted MRI from 117 patients (median age, 64 years; interquartile range [IQR], 55-73 years; 76 men) included within the Multimodal Brain Tumor Image Segmentation (BraTS) dataset plus a clinical dataset (2012-2013) with similar imaging modalities of 634 patients (median age, 59 years; IQR, 49-69 years; 382 men) with glioblastoma from six hospitals were used. Expert tumor delineations on the postcontrast images were available, but for various clinical datasets, one or more sequences were missing. The convolutional neural network, DeepMedic, was trained on combinations of complete and incomplete data with and without site-specific data. Sparsified training was introduced, which randomly simulated missing sequences during training. The effects of sparsified training and center-specific training were tested using Wilcoxon signed rank tests for paired measurements.

Results : A model trained exclusively on BraTS data reached a median Dice score of 0.81 for segmentation on BraTS test data but only 0.49 on the clinical data. Sparsified training improved performance (adjusted P < .05), even when excluding test data with missing sequences, to median Dice score of 0.67. Inclusion of site-specific data during sparsified training led to higher model performance Dice scores greater than 0.8, on par with a model based on all complete and incomplete data. For the model using BraTS and clinical training data, inclusion of site-specific data or sparsified training was of no consequence.

Conclusion : Accurate and automatic segmentation of glioblastoma on clinical scans is feasible using a model based on large, heterogeneous, and partially incomplete datasets. Sparsified training may boost the performance of a smaller model based on public and site-specific data.Supplemental material is available for this article.Published under a CC BY 4.0 license.

Eijgelaar Roelant S, Visser Martin, Müller Domenique M J, Barkhof Frederik, Vrenken Hugo, van Herk Marcel, Bello Lorenzo, Conti Nibali Marco, Rossi Marco, Sciortino Tommaso, Berger Mitchel S, Hervey-Jumper Shawn, Kiesel Barbara, Widhalm Georg, Furtner Julia, Robe Pierre A J T, Mandonnet Emmanuel, De Witt Hamer Philip C, de Munck Jan C, Witte Marnix G

2020-Sep

Radiology Radiology

Rethinking Greulich and Pyle: A Deep Learning Approach to Pediatric Bone Age Assessment Using Pediatric Trauma Hand Radiographs.

In Radiology. Artificial intelligence

Purpose : To develop a deep learning approach to bone age assessment based on a training set of developmentally normal pediatric hand radiographs and to compare this approach with automated and manual bone age assessment methods based on Greulich and Pyle (GP).

Methods : In this retrospective study, a convolutional neural network (trauma hand radiograph-trained deep learning bone age assessment method [TDL-BAAM]) was trained on 15 129 frontal view pediatric trauma hand radiographs obtained between December 14, 2009, and May 31, 2017, from Children's Hospital of New York, to predict chronological age. A total of 214 trauma hand radiographs from Hasbro Children's Hospital were used as an independent test set. The test set was rated by the TDL-BAAM model as well as a GP-based deep learning model (GPDL-BAAM) and two pediatric radiologists (radiologists 1 and 2) using the GP method. All ratings were compared with chronological age using mean absolute error (MAE), and standard concordance analyses were performed.

Results : The MAE of the TDL-BAAM model was 11.1 months, compared with 12.9 months for GPDL-BAAM (P = .0005), 14.6 months for radiologist 1 (P < .0001), and 16.0 for radiologist 2 (P < .0001). For TDL-BAAM, 95.3% of predictions were within 24 months of chronological age compared with 91.6% for GPDL-BAAM (P = .096), 86.0% for radiologist 1 (P < .0001), and 84.6% for radiologist 2 (P < .0001). Concordance was high between all methods and chronological age (intraclass coefficient > 0.93). Deep learning models demonstrated a systematic bias with a tendency to overpredict age for younger children versus radiologists who showed a consistent mean bias.

Conclusion : A deep learning model trained on pediatric trauma hand radiographs is on par with automated and manual GP-based methods for bone age assessment and provides a foundation for developing population-specific deep learning algorithms for bone age assessment in modern pediatric populations.Supplemental material is available for this article.© RSNA, 2020See also the commentary by Halabi in this issue.

Pan Ian, Baird Grayson L, Mutasa Simukayi, Merck Derek, Ruzal-Shapiro Carrie, Swenson David W, Ayyala Rama S

2020-Jul

Radiology Radiology

MRI Manufacturer Shift and Adaptation: Increasing the Generalizability of Deep Learning Segmentation for MR Images Acquired with Different Scanners.

In Radiology. Artificial intelligence

Purpose : To quantitatively evaluate the generalizability of a deep learning segmentation tool to MRI data from scanners of different MRI manufacturers and to improve the cross-manufacturer performance by using a manufacturer-adaptation strategy.

Materials and Methods : This retrospective study included 150 cine MRI datasets from three MRI manufacturers, acquired between 2017 and 2018 (n = 50 for manufacturer 1, manufacturer 2, and manufacturer 3). Three convolutional neural networks (CNNs) were trained to segment the left ventricle (LV), using datasets exclusively from images from a single manufacturer. A generative adversarial network (GAN) was trained to adapt the input image before segmentation. The LV segmentation performance, end-diastolic volume (EDV), end-systolic volume (ESV), LV mass, and LV ejection fraction (LVEF) were evaluated before and after manufacturer adaptation. Paired Wilcoxon signed rank tests were performed.

Results : The segmentation CNNs exhibited a significant performance drop when applied to datasets from different manufacturers (Dice reduced from 89.7% ± 2.3 [standard deviation] to 68.7% ± 10.8, P < .05, from 90.6% ± 2.1 to 59.5% ± 13.3, P < .05, from 89.2% ± 2.3 to 64.1% ± 12.0, P < .05, for manufacturer 1, 2, and 3, respectively). After manufacturer adaptation, the segmentation performance was significantly improved (from 68.7% ± 10.8 to 84.3% ± 6.2, P < .05, from 72.4% ± 10.2 to 85.7% ± 6.5, P < .05, for manufacturer 2 and 3, respectively). Quantitative LV function parameters were also significantly improved. For LVEF, the manufacturer adaptation increased the Pearson correlation from 0.005 to 0.89 for manufacturer 2 and from 0.77 to 0.94 for manufacturer 3.

Conclusion : A segmentation CNN well trained on datasets from one MRI manufacturer may not generalize well to datasets from other manufacturers. The proposed manufacturer adaptation can largely improve the generalizability of a deep learning segmentation tool without additional annotation.Supplemental material is available for this article.© RSNA, 2020.

Yan Wenjun, Huang Lu, Xia Liming, Gu Shengjia, Yan Fuhua, Wang Yuanyuan, Tao Qian

2020-Jul

Radiology Radiology

Cerebral Artery and Vein Segmentation in Four-dimensional CT Angiography Using Convolutional Neural Networks.

In Radiology. Artificial intelligence

Purpose : To implement and test a deep learning approach for the segmentation of the arterial and venous cerebral vasculature with four-dimensional (4D) CT angiography.

Materials and Methods : Patients who had undergone 4D CT angiography for the suspicion of acute ischemic stroke were retrospectively identified. A total of 390 patients evaluated in 2014 (n = 113) or 2018 (n = 277) were included in this study, with each patient having undergone one 4D CT angiographic scan. One hundred patients from 2014 were randomly selected, and the arteries and veins on their CT scans were manually annotated by five experienced observers. The weighted temporal average and weighted temporal variance from 4D CT angiography were used as input for a three-dimensional Dense-U-Net. The network was trained with the fully annotated cerebral vessel artery-vein maps from 60 patients. Forty patients were used for quantitative evaluation. The relative absolute volume difference and the Dice similarity coefficient are reported. The neural network segmentations from 277 patients who underwent scanning in 2018 were qualitatively evaluated by an experienced neuroradiologist using a five-point scale.

Results : The average time for processing arterial and venous cerebral vasculature with the network was less than 90 seconds. The mean Dice similarity coefficient in the test set was 0.80 ± 0.04 (standard deviation) for the arteries and 0.88 ± 0.03 for the veins. The mean relative absolute volume difference was 7.3% ± 5.7 for the arteries and 8.5% ± 4.8 for the veins. Most of the segmentations (n = 273, 99.3%) were rated as very good to perfect.

Conclusion : The proposed convolutional neural network enables accurate artery and vein segmentation with 4D CT angiography with a processing time of less than 90 seconds.© RSNA, 2020.

Meijs Midas, Pegge Sjoert A H, Vos Maria H E, Patel Ajay, van de Leemput Sil C, Koschmieder Kevin, Prokop Mathias, Meijer Frederick J A, Manniesing Rashindra

2020-Jul

Public Health Public Health

An automated framework for image classification and segmentation of fetal ultrasound images for gestational age estimation.

In Proceedings of SPIE--the International Society for Optical Engineering

Accurate assessment of fetal gestational age (GA) is critical to the clinical management of pregnancy. Industrialized countries rely upon obstetric ultrasound (US) to make this estimate. In low- and middle- income countries, automatic measurement of fetal structures using a low-cost obstetric US may assist in establishing GA without the need for skilled sonographers. In this report, we leverage a large database of obstetric US images acquired, stored and annotated by expert sonographers to train algorithms to classify, segment, and measure several fetal structures: biparietal diameter (BPD), head circumference (HC), crown rump length (CRL), abdominal circumference (AC), and femur length (FL). We present a technique for generating raw images suitable for model training by removing caliper and text annotation and describe a fully automated pipeline for image classification, segmentation, and structure measurement to estimate the GA. The resulting framework achieves an average accuracy of 93% in classification tasks, a mean Intersection over Union accuracy of 0.91 during segmentation tasks, and a mean measurement error of 1.89 centimeters, finally leading to a 1.4 day mean average error in the predicted GA compared to expert sonographer GA estimate using the Hadlock equation.

Prieto Juan C, Shah Hina, Rosenbaum Alan J, Jiang Xiaoning, Musonda Patrick, Price Joan T, Stringer Elizabeth M, Vwalika Bellington, Stamilio David M, Stringer Jeffrey S A

2021-Feb

Fetal ultrasound, GA estimation, Machine learning

General General

A COVID-19 time series forecasting model based on MLP ANN.

In Procedia computer science

With the accelerated spread of COVID-19 worldwide and its potentially fatal effects on human health, the development of a tool that effectively describes and predicts the number of infected cases and deaths over time becomes relevant. This makes it possible for administrative sectors and the population itself to become aware and act more precisely. In this work, a machine learning model based on the multilayer Perceptron artificial neural network structure was used, which effectively predicts the behavior of the series mentioned in up to six days. The model, which is trained with data from 30 countries together in a 20-day context, is assessed using global and local MSE and MAE measures. For the construction of training and test sets, four time series (number of: accumulated infected cases, new cases, accumulated deaths and new deaths) from each country are used, which are started on the day of the first confirmed infection case. In order to soften the sudden transitions between samples, a moving average filter with a window size 3 and a normalization by maximum value were used. It is intended to make the model's predictions available online, collaborating with the fight against the pandemic.

Borghi Pedro Henrique, Zakordonets Oleksandr, Teixeira João Paulo

2021

COVID-19 Brazil forecast, COVID-19 Italy forecast, COVID-19 worldwide forecast

General General

Supervised multi-specialist topic model with applications on large-scale electronic health record data

ArXiv Preprint

Motivation: Electronic health record (EHR) data provides a new venue to elucidate disease comorbidities and latent phenotypes for precision medicine. To fully exploit its potential, a realistic data generative process of the EHR data needs to be modelled. We present MixEHR-S to jointly infer specialist-disease topics from the EHR data. As the key contribution, we model the specialist assignments and ICD-coded diagnoses as the latent topics based on patient's underlying disease topic mixture in a novel unified supervised hierarchical Bayesian topic model. For efficient inference, we developed a closed-form collapsed variational inference algorithm to learn the model distributions of MixEHR-S. We applied MixEHR-S to two independent large-scale EHR databases in Quebec with three targeted applications: (1) Congenital Heart Disease (CHD) diagnostic prediction among 154,775 patients; (2) Chronic obstructive pulmonary disease (COPD) diagnostic prediction among 73,791 patients; (3) future insulin treatment prediction among 78,712 patients diagnosed with diabetes as a mean to assess the disease exacerbation. In all three applications, MixEHR-S conferred clinically meaningful latent topics among the most predictive latent topics and achieved superior target prediction accuracy compared to the existing methods, providing opportunities for prioritizing high-risk patients for healthcare services. MixEHR-S source code and scripts of the experiments are freely available at https://github.com/li-lab-mcgill/mixehrS

Ziyang Song, Xavier Sumba Toral, Yixin Xu, Aihua Liu, Liming Guo, Guido Powell, Aman Verma, David Buckeridge, Ariane Marelli, Yue Li

2021-05-04

General General

Fully Automatic Volume Measurement of the Spleen at CT Using Deep Learning.

In Radiology. Artificial intelligence

Purpose : To develop a fully automated algorithm for spleen segmentation and to assess the performance of this algorithm in a large dataset.

Materials and Methods : In this retrospective study, a three-dimensional deep learning network was developed to segment the spleen on thorax-abdomen CT scans. Scans were extracted from patients undergoing oncologic treatment from 2014 to 2017. A total of 1100 scans from 1100 patients were used in this study, and 400 were selected for development of the algorithm. For testing, a dataset of 50 scans was annotated to assess the segmentation accuracy and was compared against the splenic index equation. In a qualitative observer experiment, an enriched set of 100 scan-pairs was used to evaluate whether the algorithm could aid a radiologist in assessing splenic volume change. The reference standard was set by the consensus of two other independent radiologists. A Mann-Whitney U test was conducted to test whether there was a performance difference between the algorithm and the independent observer.

Results : The algorithm and the independent observer obtained comparable Dice scores (P = .834) on the test set of 50 scans of 0.962 and 0.964, respectively. The radiologist had an agreement with the reference standard in 81% (81 of 100) of the cases after a visual classification of volume change, which increased to 92% (92 of 100) when aided by the algorithm.

Conclusion : A segmentation method based on deep learning can accurately segment the spleen on CT scans and may help radiologists to detect abnormal splenic volumes and splenic volume changes.Supplemental material is available for this article.© RSNA, 2020.

Humpire-Mamani Gabriel E, Bukala Joris, Scholten Ernst T, Prokop Mathias, van Ginneken Bram, Jacobs Colin

2020-Jul

Radiology Radiology

Deep Learning-based Approach for Automated Assessment of Interstitial Lung Disease in Systemic Sclerosis on CT Images.

In Radiology. Artificial intelligence

Purpose : To develop a deep learning algorithm for the automatic assessment of the extent of systemic sclerosis (SSc)-related interstitial lung disease (ILD) on chest CT images.

Materials and Methods : This retrospective study included 208 patients with SSc (median age, 57 years; 167 women) evaluated between January 2009 and October 2017. A multicomponent deep neural network (AtlasNet) was trained on 6888 fully annotated CT images (80% for training and 20% for validation) from 17 patients with no, mild, or severe lung disease. The model was tested on a dataset of 400 images from another 20 patients, independently partially annotated by three radiologist readers. The ILD contours from the three readers and the deep learning neural network were compared by using the Dice similarity coefficient (DSC). The correlation between disease extent obtained from the deep learning algorithm and that obtained by using pulmonary function tests (PFTs) was then evaluated in the remaining 171 patients and in an external validation dataset of 31 patients based on the analysis of all slices of the chest CT scan. The Spearman rank correlation coefficient (ρ) was calculated to evaluate the correlation between disease extent and PFT results.

Results : The median DSCs between the readers and the deep learning ILD contours ranged from 0.74 to 0.75, whereas the median DSCs between contours from radiologists ranged from 0.68 to 0.71. The disease extent obtained from the algorithm, by analyzing the whole CT scan, correlated with the diffusion lung capacity for carbon monoxide, total lung capacity, and forced vital capacity (ρ = -0.76, -0.70, and -0.62, respectively; P < .001 for all) in the dataset for the correlation with PFT results. The disease extents correlated with diffusion lung capacity for carbon monoxide, total lung capacity, and forced vital capacity were ρ = -0.65, -0.70, and -0.57, respectively, in the external validation dataset (P < .001 for all).

Conclusion : The developed algorithm performed similarly to radiologists for disease-extent contouring, which correlated with pulmonary function to assess CT images from patients with SSc-related ILD.Supplemental material is available for this article.© RSNA, 2020.

Chassagnon Guillaume, Vakalopoulou Maria, Régent Alexis, Zacharaki Evangelia I, Aviram Galit, Martin Charlotte, Marini Rafael, Bus Norbert, Jerjir Naïm, Mekinian Arsène, Hua-Huy Thông, Monnier-Cholley Laurence, Benmostefa Nouria, Mouthon Luc, Dinh-Xuan Anh-Tuan, Paragios Nikos, Revel Marie-Pierre

2020-Jul

Radiology Radiology

Construction of a Machine Learning Dataset through Collaboration: The RSNA 2019 Brain CT Hemorrhage Challenge.

In Radiology. Artificial intelligence

This dataset is composed of annotations of the five hemorrhage subtypes (subarachnoid, intraventricular, subdural, epidural, and intraparenchymal hemorrhage) typically encountered at brain CT.

Flanders Adam E, Prevedello Luciano M, Shih George, Halabi Safwan S, Kalpathy-Cramer Jayashree, Ball Robyn, Mongan John T, Stein Anouk, Kitamura Felipe C, Lungren Matthew P, Choudhary Gagandeep, Cala Lesley, Coelho Luiz, Mogensen Monique, Morón Fanny, Miller Elka, Ikuta Ichiro, Zohrabian Vahe, McDonnell Olivia, Lincoln Christie, Shah Lubdha, Joyner David, Agarwal Amit, Lee Ryan K, Nath Jaya

2020-May

Radiology Radiology

Deep Learning to Automate Reference-Free Image Quality Assessment of Whole-Heart MR Images.

In Radiology. Artificial intelligence

Purpose : To develop and characterize an algorithm that mimics human expert visual assessment to quantitatively determine the quality of three-dimensional (3D) whole-heart MR images.

Materials and Methods : In this study, 3D whole-heart cardiac MRI scans from 424 participants (average age, 57 years ± 18 [standard deviation]; 66.5% men) were used to generate an image quality assessment algorithm. A deep convolutional neural network for image quality assessment (IQ-DCNN) was designed, trained, optimized, and cross-validated on a clinical database of 324 (training set) scans. On a separate test set (100 scans), two hypotheses were tested: (a) that the algorithm can assess image quality in concordance with human expert assessment as assessed by human-machine correlation and intra- and interobserver agreement and (b) that the IQ-DCNN algorithm may be used to monitor a compressed sensing reconstruction process where image quality progressively improves. Weighted κ values, agreement and disagreement counts, and Krippendorff α reliability coefficients were reported.

Results : Regression performance of the IQ-DCNN was within the range of human intra- and interobserver agreement and in very good agreement with the human expert (R2 = 0.78, κ = 0.67). The image quality assessment during compressed sensing reconstruction correlated with the cost function at each iteration and was successfully applied to rank the results in very good agreement with the human expert.

Conclusion : The proposed IQ-DCNN was trained to mimic expert visual image quality assessment of 3D whole-heart MR images. The results from the IQ-DCNN were in good agreement with human expert reading, and the network was capable of automatically comparing different reconstructed volumes.Supplemental material is available for this article.© RSNA, 2020.

Piccini Davide, Demesmaeker Robin, Heerfordt John, Yerly Jérôme, Di Sopra Lorenzo, Masci Pier Giorgio, Schwitter Juerg, Van De Ville Dimitri, Richiardi Jonas, Kober Tobias, Stuber Matthias

2020-May

Radiology Radiology

The Case for User-Centered Artificial Intelligence in Radiology.

In Radiology. Artificial intelligence

Past technology transition successes and failures have demonstrated the importance of user-centered design and the science of human factors; these approaches will be critical to the success of artificial intelligence in radiology.

Filice Ross W, Ratwani Raj M

2020-May

Radiology Radiology

Generative Adversarial Networks Improve the Reproducibility and Discriminative Power of Radiomic Features.

In Radiology. Artificial intelligence

Purpose : To assess the contribution of a generative adversarial network (GAN) to improve intermanufacturer reproducibility of radiomic features (RFs).

Materials and Methods : The authors retrospectively developed a cycle-GAN to translate texture information from chest radiographs acquired using one manufacturer (Siemens) to chest radiographs acquired using another (Philips), producing fake chest radiographs with different textures. The authors prospectively evaluated the ability of this texture-translation cycle-GAN to reduce the intermanufacturer variability of RFs extracted from the lung parenchyma. This study assessed the cycle-GAN's ability to fool several machine learning (ML) classifiers tasked with recognizing the manufacturer on the basis of chest radiography inputs. The authors also evaluated the cycle-GAN's ability to mislead radiologists who were asked to perform the same recognition task. Finally, the authors tested whether the cycle-GAN had an impact on radiomic diagnostic accuracy for chest radiography in patients with congestive heart failure (CHF).

Results : RFs, extracted from chest radiographs after the cycle-GAN's texture translation (fake chest radiographs), showed decreased intermanufacturer RF variability. Using cycle-GAN-generated chest radiographs as inputs, ML classifiers categorized the fake chest radiographs as belonging to the target manufacturer rather than to a native one. Moreover, cycle-GAN fooled two experienced radiologists who identified fake chest radiographs as belonging to a target manufacturer class. Finally, reducing intermanufacturer RF variability with cycle-GAN improved the discriminative power of RFs for patients without CHF versus patients with CHF (from 55% to 73.5%, P < .001).

Conclusion : Both ML classifiers and radiologists had difficulty recognizing the chest radiographs' manufacturer. The cycle-GAN improved RF intermanufacturer reproducibility and discriminative power for identifying patients with CHF. This deep learning approach may help counteract the sensitivity of RFs to differences in acquisition.Supplemental material is available for this article.© RSNA, 2020See also the commentary by Alderson in this issue.

Marcadent Sandra, Hofmeister Jeremy, Preti Maria Giulia, Martin Steve P, Van De Ville Dimitri, Montet Xavier

2020-May

Surgery Surgery

Convolutional Neural Networks for Automatic Risser Stage Assessment.

In Radiology. Artificial intelligence

Purpose : To develop an automatic method for the assessment of the Risser stage using deep learning that could be used in the management panel of adolescent idiopathic scoliosis (AIS).

Materials and Methods : In this institutional review board approved-study, a total of 1830 posteroanterior radiographs of patients with AIS (age range, 10-18 years, 70% female) were collected retrospectively and graded manually by six trained readers using the United States Risser staging system. Each radiograph was preprocessed and cropped to include the entire pelvic region. A convolutional neural network was trained to automatically grade conventional radiographs according to the Risser classification. The network was then validated by comparing its accuracy against the interobserver variability of six trained graders from the authors' institution using the Fleiss κ statistical measure.

Results : Overall agreement between the six observers was fair, with a κ coefficient of 0.65 for the experienced graders and agreement of 74.5%. The automatic grading method obtained a κ coefficient of 0.72, which is a substantial agreement with the ground truth, and an overall accuracy of 78.0%.

Conclusion : The high accuracy of the model presented here compared with human readers suggests that this work may provide a new method for standardization of Risser grading. The model could assist physicians with the task, as well as provide additional insights in the assessment of bone maturity based on radiographs.© RSNA, 2020.

Kaddioui Houda, Duong Luc, Joncas Julie, Bellefleur Christian, Nahle Imad, Chémaly Olivier, Nault Marie-Lyne, Parent Stefan, Grimard Guy, Labelle Hubert

2020-May

General General

Predicting Admissions From a Paediatric Emergency Department - Protocol for Developing and Validating a Low-Dimensional Machine Learning Prediction Model.

In Frontiers in big data

Introduction: Patients boarding in the Emergency Department can contribute to overcrowding, leading to longer waiting times and patients leaving without being seen or completing their treatment. The early identification of potential admissions could act as an additional decision support tool to alert clinicians that a patient needs to be reviewed for admission and would also be of benefit to bed managers in advance bed planning for the patient. We aim to create a low-dimensional model predicting admissions early from the paediatric Emergency Department. Methods and Analysis: The methodology Cross Industry Standard Process for Data Mining (CRISP-DM) will be followed. The dataset will comprise of 2 years of data, ~76,000 records. Potential predictors were identified from previous research, comprising of demographics, registration details, triage assessment, hospital usage and past medical history. Fifteen models will be developed comprised of 3 machine learning algorithms (Logistic regression, naïve Bayes and gradient boosting machine) and 5 sampling methods, 4 of which are aimed at addressing class imbalance (undersampling, oversampling, and synthetic oversampling techniques). The variables of importance will then be identified from the optimal model (selected based on the highest Area under the curve) and used to develop an additional low-dimensional model for deployment. Discussion: A low-dimensional model comprised of routinely collected data, captured up to post triage assessment would benefit many hospitals without data rich platforms for the development of models with a high number of predictors. Novel to the planned study is the use of data from the Republic of Ireland and the application of sampling techniques aimed at improving model performance impacted by an imbalance between admissions and discharges in the outcome variable.

Leonard Fiona, Gilligan John, Barrett Michael J

2021

admission, emergency department, machine learning, paediatric, prediction, protocol

General General

Corrigendum: The Autonomous Mind: The Right to Freedom of Thought in the Twenty-First Century.

In Frontiers in artificial intelligence

[This corrects the article DOI: 10.3389/frai.2019.00019.].

McCarthy-Jones Simon

2021

big data, human rights, law, machine learning, privacy, psychology

General General

Machine Learning Analysis of Hydrologic Exchange Flows and Transit Time Distributions in a Large Regulated River.

In Frontiers in artificial intelligence

Hydrologic exchange between river channels and adjacent subsurface environments is a key process that influences water quality and ecosystem function in river corridors. High-resolution numerical models were often used to resolve the spatial and temporal variations of exchange flows, which are computationally expensive. In this study, we adopt Random Forest (RF) and Extreme Gradient Boosting (XGB) approaches for deriving reduced order models of hydrologic exchange flows and associated transit time distributions, with integrated field observations (e.g., bathymetry) and hydrodynamic simulation data (e.g., river velocity, depth). The setup allows an improved understanding of the influences of various physical, spatial, and temporal factors on the hydrologic exchange flows and transit times. The predictors also contain those derived using hybrid clustering, leveraging our previous work on river corridor system hydromorphic classification. The machine learning-based predictive models are developed and validated along the Columbia River Corridor, and the results show that the top parameters are the thickness of the top geological formation layer, the flow regime, river velocity, and river depth; the RF and XGB models can achieve 70% to 80% accuracy and therefore are effective alternatives to the computational demanding numerical models of exchange flows and transit time distributions. Each machine learning model with its favorable configuration and setup have been evaluated. The transferability of the models to other river reaches and larger scales, which mostly depends on data availability, is also discussed.

Ren Huiying, Song Xuehang, Fang Yilin, Hou Z Jason, Scheibe Timothy D

2021

extreme gradient boosting, hydrologic exchange flows, machine learning, random forest, spatial heterogeneity, transit time

General General

Total variation with modified group sparsity for CT reconstruction under low SNR.

In Journal of X-ray science and technology

BACKGROUND AND OBJECTIVE : Since the stair artifacts may affect non-destructive testing (NDT) and diagnosis in the later stage, an applicable model is desperately needed, which can deal with the stair artifacts and preserve the edges. However, the classical total variation (TV) algorithm only considers the sparsity of the gradient transformed image. The objective of this study is to introduce and test a new method based on group sparsity to address the low signal-to-noise ratio (SNR) problem.

METHODS : This study proposes a weighted total variation with overlapping group sparsity model. This model combines the Gaussian kernel and overlapping group sparsity into TV model denoted as GOGS-TV, which considers the structure sparsity of the image to be reconstructed to deal with the stair artifacts. On one hand, TV is the accepted commercial algorithm, and it can work well in many situations. On the other hand, the Gaussian kernel can associate the points around each pixel. Quantitative assessments are implemented to verify this merit.

RESULTS : Numerical simulations are performed to validate the presented method, compared with the classical simultaneous algebraic reconstruction technique (SART) and the state-of-the-art TV algorithm. It confirms the significantly improved SNR of the reconstruction images both in suppressing the noise and preserving the edges using new GOGS-TV model.

CONCLUSIONS : The proposed GOGS-TV model demonstrates its advantages to reduce stair artifacts especially in low SNR reconstruction because this new model considers both the sparsity of the gradient image and the structured sparsity. Meanwhile, the Gaussian kernel is utilized as a weighted factor that can be adapted to the global distribution.

Zhang Lingli

2021-Apr-26

Computed tomography (CT), group sparsity, image reconstruction, signal-to-noise ratio (SNR)., total variation (TV)

General General

The Role of Artificial Intelligence in Fighting the COVID-19 Pandemic.

In Information systems frontiers : a journal of research and innovation

The first few months of 2020 have profoundly changed the way we live our lives and carry out our daily activities. Although the widespread use of futuristic robotaxis and self-driving commercial vehicles has not yet become a reality, the COVID-19 pandemic has dramatically accelerated the adoption of Artificial Intelligence (AI) in different fields. We have witnessed the equivalent of two years of digital transformation compressed into just a few months. Whether it is in tracing epidemiological peaks or in transacting contactless payments, the impact of these developments has been almost immediate, and a window has opened up on what is to come. Here we analyze and discuss how AI can support us in facing the ongoing pandemic. Despite the numerous and undeniable contributions of AI, clinical trials and human skills are still required. Even if different strategies have been developed in different states worldwide, the fight against the pandemic seems to have found everywhere a valuable ally in AI, a global and open-source tool capable of providing assistance in this health emergency. A careful AI application would enable us to operate within this complex scenario involving healthcare, society and research.

Piccialli Francesco, di Cola Vincenzo Schiano, Giampaolo Fabio, Cuomo Salvatore

2021-Apr-26

Artificial intelligence, COVID-19, Deep learning, Healthcare, Machine learning, Review, SARS-CoV-2, Survey

Pathology Pathology

Remote Pathological Gait Classification System

ArXiv Preprint

Several pathologies can alter the way people walk, i.e. their gait. Gait analysis can therefore be used to detect impairments and help diagnose illnesses and assess patient recovery. Using vision-based systems, diagnoses could be done at home or in a clinic, with the needed computation being done remotely. State-of-the-art vision-based gait analysis systems use deep learning, requiring large datasets for training. However, to our best knowledge, the biggest publicly available pathological gait dataset contains only 10 subjects, simulating 4 gait pathologies. This paper presents a new dataset called GAIT-IT, captured from 21 subjects simulating 4 gait pathologies, with 2 severity levels, besides normal gait, being considerably larger than publicly available gait pathology datasets, allowing to train a deep learning model for gait pathology classification. Moreover, it was recorded in a professional studio, making it possible to obtain nearly perfect silhouettes, free of segmentation errors. Recognizing the importance of remote healthcare, this paper proposes a prototype of a web application allowing to upload a walking person's video, possibly acquired using a smartphone camera, and execute a web service that classifies the person's gait as normal or across different pathologies. The web application has a user friendly interface and could be used by healthcare professionals or other end users. An automatic gait analysis system is also developed and integrated with the web application for pathology classification. Compared to state-of-the-art solutions, it achieves a drastic reduction in the number of model parameters, which means significantly lower memory requirements, as well as lower training and execution times. Classification accuracy is on par with the state-of-the-art.

Pedro Albuquerque, Joao Machado, Tanmay Tulsidas Verlekar, Luis Ducla Soares, Paulo Lobato Correia

2021-05-04

General General

The Promise of AI for DILI Prediction.

In Frontiers in artificial intelligence

Drug-induced liver injury (DILI) is a common reason for the withdrawal of a drug from the market. Early assessment of DILI risk is an essential part of drug development, but it is rendered challenging prior to clinical trials by the complex factors that give rise to liver damage. Artificial intelligence (AI) approaches, particularly those building on machine learning, range from random forests to more recent techniques such as deep learning, and provide tools that can analyze chemical compounds and accurately predict some of their properties based purely on their structure. This article reviews existing AI approaches to predicting DILI and elaborates on the challenges that arise from the as yet limited availability of data. Future directions are discussed focusing on rich data modalities, such as 3D spheroids, and the slow but steady increase in drugs annotated with DILI risk labels.

Vall Andreu, Sabnis Yogesh, Shi Jiye, Class Reiner, Hochreiter Sepp, Klambauer Günter

2021

artificial intelligence, deep learning, drug-induced liver injury, machine learning, neural networks

oncology Oncology

Artificial Intelligence for Prognostic Scores in Oncology: a Benchmarking Study.

In Frontiers in artificial intelligence

Introduction: Prognostic scores are important tools in oncology to facilitate clinical decision-making based on patient characteristics. To date, classic survival analysis using Cox proportional hazards regression has been employed in the development of these prognostic scores. With the advance of analytical models, this study aimed to determine if more complex machine-learning algorithms could outperform classical survival analysis methods. Methods: In this benchmarking study, two datasets were used to develop and compare different prognostic models for overall survival in pan-cancer populations: a nationwide EHR-derived de-identified database for training and in-sample testing and the OAK (phase III clinical trial) dataset for out-of-sample testing. A real-world database comprised 136K first-line treated cancer patients across multiple cancer types and was split into a 90% training and 10% testing dataset, respectively. The OAK dataset comprised 1,187 patients diagnosed with non-small cell lung cancer. To assess the effect of the covariate number on prognostic performance, we formed three feature sets with 27, 44 and 88 covariates. In terms of methods, we benchmarked ROPRO, a prognostic score based on the Cox model, against eight complex machine-learning models: regularized Cox, Random Survival Forests (RSF), Gradient Boosting (GB), DeepSurv (DS), Autoencoder (AE) and Super Learner (SL). The C-index was used as the performance metric to compare different models. Results: For in-sample testing on the real-world database the resulting C-index [95% CI] values for RSF 0.720 [0.716, 0.725], GB 0.722 [0.718, 0.727], DS 0.721 [0.717, 0.726] and lastly, SL 0.723 [0.718, 0.728] showed significantly better performance as compared to ROPRO 0.701 [0.696, 0.706]. Similar results were derived across all feature sets. However, for the out-of-sample validation on OAK, the stronger performance of the more complex models was not apparent anymore. Consistently, the increase in the number of prognostic covariates did not lead to an increase in model performance. Discussion: The stronger performance of the more complex models did not generalize when applied to an out-of-sample dataset. We hypothesize that future research may benefit by adding multimodal data to exploit advantages of more complex models.

Loureiro Hugo, Becker Tim, Bauer-Mehren Anna, Ahmidi Narges, Weberpals Janick

2021

electronic health records, machine learning, prognostic scores, real world data, survival analyisis

General General

An Application of an Embedded Model Estimator to a Synthetic Nonstationary Reservoir Model With Multiple Secondary Variables.

In Frontiers in artificial intelligence

A method (Ember) for nonstationary spatial modeling with multiple secondary variables by combining Geostatistics with Random Forests is applied to a three-dimensional Reservoir Model. It extends the Random Forest method to an interpolation algorithm retaining similar consistency properties to both Geostatistical algorithms and Random Forests. It allows embedding of simpler interpolation algorithms into the process, combining them through the Random Forest training process. The algorithm estimates a conditional distribution at each target location. The family of such distributions is called the model envelope. An algorithm to produce stochastic simulations from the envelope is demonstrated. This algorithm allows the influence of the secondary variables, as well as the variability of the result to vary by location in the simulation.

Daly Colin

2021

Ember, geostatistics, machine learning, random forest, reservoir modeling, spatial statistics

Cardiology Cardiology

AI-Based Estimation of End-Systolic Elastance From Arm-Pressure and Systolic Time Intervals.

In Frontiers in artificial intelligence

Left ventricular end-systolic elastance (Ees) is a major determinant of cardiac systolic function and ventricular-arterial interaction. Previous methods for the Ees estimation require the use of the echocardiographic ejection fraction (EF). However, given that EF expresses the stroke volume as a fraction of end-diastolic volume (EDV), accurate interpretation of EF is attainable only with the additional measurement of EDV. Hence, there is still need for a simple, reliable, noninvasive method to estimate Ees. This study proposes a novel artificial intelligence-based approach to estimate Ees using the information embedded in clinically relevant systolic time intervals, namely the pre-ejection period (PEP) and ejection time (ET). We developed a training/testing scheme using virtual subjects (n = 4,645) from a previously validated in-silico model. Extreme Gradient Boosting regressor was employed to model Ees using as inputs arm cuff pressure, PEP, and ET. Results showed that Ees can be predicted with high accuracy achieving a normalized RMSE equal to 9.15% (r = 0.92) for a wide range of Ees values from 1.2 to 4.5 mmHg/ml. The proposed model was found to be less sensitive to measurement errors (±10-30% of the actual value) in blood pressure, presenting low test errors for the different levels of noise (RMSE did not exceed 0.32 mmHg/ml). In contrast, a high sensitivity was reported for measurements errors in the systolic timing features. It was demonstrated that Ees can be reliably estimated from the traditional arm-pressure and echocardiographic PEP and ET. This approach constitutes a step towards the development of an easy and clinically applicable method for assessing left ventricular systolic function.

Bikia Vasiliki, Adamopoulos Dionysios, Pagoulatou Stamatia, Rovas Georgios, Stergiopulos Nikolaos

2021

cardiac monitoring, contractility, heart, noninvasive, regression analysis

Public Health Public Health

Experimental and natural evidence of SARS-CoV-2 infection-induced activation of type I interferon responses.

In iScience

Type I interferons (IFNs) are our first line of defence against virus infection. Recent studies have suggested the ability of SARS-CoV-2 proteins to inhibit IFN responses. Emerging data also suggest that timing and extent of IFN production is associated with manifestation of COVID-19 severity. In spite of progress in understanding how SARS-CoV-2 activates antiviral responses, mechanistic studies into wildtype SARS-CoV-2-mediated induction and inhibition of human type I IFN responses are scarce. Here we demonstrate that SARS-CoV-2 infection induces a type I IFN response in vitro and in moderate cases of COVID-19. In vitro stimulation of type I IFN expression and signaling in human airway epithelial cells is associated with activation of canonical transcriptions factors, and SARS-CoV-2 is unable to inhibit exogenous induction of these responses. Furthermore, we show that physiological levels of IFNα detected in patients with moderate COVID-19 is sufficient to suppress SARS-CoV-2 replication in human airway cells.

Banerjee Arinjay, El-Sayes Nader, Budylowski Patrick, Jacob Rajesh Abraham, Richard Daniel, Maan Hassaan, Aguiar Jennifer A, Demian Wael L, Baid Kaushal, D’Agostino Michael R, Ang Jann Catherine, Murdza Tetyana, Tremblay Benjamin J-M, Afkhami Sam, Karimzadeh Mehran, Irving Aaron T, Yip Lily, Ostrowski Mario, Hirota Jeremy A, Kozak Robert, Capellini Terence D, Miller Matthew S, Wang Bo, Mubareka Samira, McGeer Allison J, McArthur Andrew G, Doxey Andrew C, Mossman Karen

2021-Apr-26

oncology Oncology

Radiomic biomarkers of tumor immune biology and immunotherapy response.

In Clinical and translational radiation oncology

Immunotherapies are leading to improved outcomes for many cancers, including those with devastating prognoses. As therapies like immune checkpoint inhibitors (ICI) become a mainstay in treatment regimens, many concurrent challenges have arisen - for instance, delineating clinical responders from non-responders. Predicting response has proven to be difficult given a lack of consistent and accurate biomarkers, heterogeneity of the tumor microenvironment (TME), and a poor understanding of resistance mechanisms. For the most part, imaging data have remained an untapped, yet abundant, resource to address these challenges. In recent years, quantitative image analyses have highlighted the utility of medical imaging in predicting tumor phenotypes, prognosis, and therapeutic response. These studies have been fueled by an explosion of resources in high-throughput mining of image features (i.e. radiomics) and artificial intelligence. In this review, we highlight current progress in radiomics to understand tumor immune biology and predict clinical responses to immunotherapies. We also discuss limitations in these studies and future directions for the field, particularly if high-dimensional imaging data are to play a larger role in precision medicine.

Wang Jarey H, Wahid Kareem A, van Dijk Lisanne V, Farahani Keyvan, Thompson Reid F, Fuller Clifton David

2021-May

Biomarkers, Imaging genomics, Immunotherapy, Precision medicine, Radiogenomics, Radiomics, Tumor immunology

Surgery Surgery

Real-time deep learning-based colorectal polyp localization on clinical video footage achievable with a wide array of hardware configurations.

In Endoscopy international open

Background and study aims  Several computer-assisted polyp detection systems have been proposed, but they have various limitations, from utilizing outdated neural network architectures to a requirement for multi-graphics processing unit (GPU) processing, to validating on small or non-robust datasets. To address these problems, we developed a system based on a state-of-the-art convolutional neural network architecture able to detect polyps in real time on a single GPU and tested on both public datasets and full clinical examination recordings. Methods  The study comprised 165 colonoscopy procedure recordings and 2678 still photos gathered retrospectively. The system was trained on 81,962 polyp frames in total and then tested on footage from 42 colonoscopies and CVC-ClinicDB, CVC-ColonDB, Hyper-Kvasir, and ETIS-Larib public datasets. Clinical videos were evaluated for polyp detection and false-positive rates whereas the public datasets were assessed for F1 score. The system was tested for runtime performance on a wide array of hardware. Results  The performance on public datasets varied from an F1 score of 0.727 to 0.942. On full examination videos, it detected 94 % of the polyps found by the endoscopist with a 3 % false-positive rate and identified additional polyps that were missed during initial video assessment. The system's runtime fits within the real-time constraints on all but one of the hardware configurations. Conclusions  We have created a polyp detection system with a post-processing pipeline that works in real time on a wide array of hardware. The system does not require extensive computational power, which could help broaden the adaptation of new commercially available systems.

Podlasek Jeremi, Heesch Mateusz, Podlasek Robert, Kilisiński Wojciech, Filip Rafał

2021-May

General General

A Novel Machine Learning Predictive Tool Assessing Outpatient or Inpatient Designation for Medicare Patients Undergoing Total Hip Arthroplasty.

In Arthroplasty today

Background : The Centers for Medicare and Medicaid Services removed total hip arthroplasty (THA) from the inpatient-only list. This has created significant confusion regarding which patients qualify for an inpatient designation. The purpose of this study is to develop and validate a novel predictive tool for preoperatively objectively determining "outpatient" vs "inpatient" status for THA in the Medicare population.

Methods : A cohort of Medicare patients undergoing primary THA between January 2017 and September 2019 was retrospectively reviewed. A machine learning model was trained using 80% of the THA patients, and the remaining 20% was used for testing the model performance in terms of accuracy and the average area under the receiver operating characteristic curve. Feature importance was obtained for each feature used in the model.

Results : One thousand ninety-one patients had outpatient stays, and 318 qualified for inpatient designation. Significant associations were demonstrated between inpatient designations and the following: higher BMI, increased patient age, better preoperative functional scores, higher American Society of Anesthesiologist Physical Status Classification, higher Modified Frailty Index, higher Charlson Comorbidity Index, female gender, and numerous comorbidities. The XGBoost model for predicting an inpatient or outpatient stay was 78.7% accurate with the area under the receiver operating characteristic curve to be 81.5%.

Conclusions : Using readily available key baseline characteristics, functional scores and comorbidities, this machine-learning model accurately predicts an "outpatient" or "inpatient" stay after THA in the Medicare population. BMI, age, functional scores, and American Society of Anesthesiologist Physical Status Classification had the highest influence on this predictive model.

Kugelman David N, Teo Greg, Huang Shengnan, Doran Michael G, Singh Vivek, Long William J

2021-Apr

Arthroplasty inpatient only, Medicare bundle payment, Medicare inpatient only list, Medicare total hip, Predictive model, Total hip arthroplasty

General General

Dataset of Pakistan Sign Language and Automatic Recognition of Hand Configuration of Urdu Alphabet through Machine Learning.

In Data in brief

Social correspondence is one of the most significant columns that the public dependent on. Notably, language is the best way to communicate and associate with one another both verbally and nonverbally. There is a persistent communication gap among deaf and non-deaf communities because non-deaf people have less understanding of sign languages. Every region/country has its sign language. In Pakistan, the sign language of Urdu is a visual gesture language that is being used for communication among deaf peoples. However, the dataset of Pakistan Sign Language (PSL) is not available publicly. The dataset of PSL has been generated by acquiring images of different hand configurations through a webcam. In this work, 40 images of each hand configuration with multiple orientations have been captured. In addition, we developed, an interactive android mobile application based on machine learning that minimized the communication barrier between the deaf and non-deaf communities by using the PSL dataset. The android application recognizes the Urdu alphabet from input hand configuration.

Imran Ali, Razzaq Abdul, Baig Irfan Ahmad, Hussain Aamir, Shahid Sharaiz, Rehman Tausif-Ur

2021-Jun

Deaf people communication, Hand configuration, Machine learning, Mobile app, Pakistan sign language

oncology Oncology

Automated fiducial marker detection and localization in volumetric computed tomography images: a three-step hybrid approach with deep learning.

In Journal of medical imaging (Bellingham, Wash.)

Purpose: Automating fiducial detection and localization in the patient's pre-operative images can lead to better registration accuracy, reduced human errors, and shorter intervention time. Most current approaches are optimized for a single marker type, mainly spherical adhesive markers. A fully automated algorithm is proposed and evaluated for screw and spherical titanium fiducials, typically used in high-accurate frameless surgical navigation. Approach: The algorithm builds on previous approaches with morphological functions and pose estimation algorithms. A 3D convolutional neural network (CNN) is proposed for the fiducial classification task and evaluated for both traditional closed-set and emerging open-set classifiers. A proposed digital ground-truth experiment, with cone-beam computed tomography (CBCT) imaging software, is performed to determine the localization accuracy of the algorithm. The localized fiducial positions in the CBCT images by the presented algorithm were compared to the actual known positions in the virtual phantom models. The difference represents the fiducial localization error (FLE). Results: A total of 241 screws, 151 spherical fiducials, and 1550 other structures are identified with the best true positive rate 95.9% for screw and 99.3% for spherical fiducials at 8.7% and 3.4% false positive rate, respectively. The best achieved FLE mean and its standard deviation for a screw and spherical marker are 58 (14) and 14 ( 6 )    μ m , respectively. Conclusions: Accurate marker detection and localization were achieved, with spherical fiducials being superior to screws. Large marker volume and smaller voxel size yield significantly smaller FLEs. Attenuating noise by mesh smoothing has a minor effect on FLE. Future work will focus on expanding the CNN for image segmentation.

Regodić Milovan, Bardosi Zoltan, Freysinger Wolfgang

2021-Mar

convolutional neural network, fiducial markers, image-guided surgery, open-set recognition, phantom models, virtual computed tomography

General General

Quantitative Gait Analysis Using a Pose-Estimation Algorithm with a Single 2D-Video of Parkinson's Disease Patients.

In Journal of Parkinson's disease

BACKGROUND : Clinician-based rating scales or questionnaires for gait in Parkinson's disease (PD) are subjective and sensor-based analysis is limited in accessibility.

OBJECTIVE : To develop an easily accessible and objective tool to evaluate gait in PD patients, we analyzed gait from a single 2-dimensional (2D) video.

METHODS : We prospectively recorded 2D videos of PD patients (n = 16) and healthy controls (n = 15) performing the timed up and go test (TUG). The gait was simultaneously evaluated with a pressure-sensor (GAITRite). We estimated the 3D position of toes and heels with a deep-learning based pose-estimation algorithm and calculated gait parameters including step length, step length variability, gait velocity and step cadence which was validated with the result from the GAITRite. We further calculated the time and steps required for turning. Then, we applied the algorithm to previously recorded and archived videos of PD patients (n = 32) performing the TUG.

RESULTS : From the validation experiment, gait parameters derived from video tracking were in excellent agreement with the parameters obtained with the GAITRite. (Intraclass correlation coefficient > 0.9). From the analysis with the archived videos, step length, gait velocity, number of steps, and the time required for turning were significantly correlated (Absolute R > 0.4, p < 0.005) with the Freezing of gait questionnaire, Unified PD Rating scale part III total score, HY stage and postural instability. Furthermore, the video-based tracking objectively measured significant improvement of step length, gait velocity, steps and the time required for turning with antiparkinsonian medication.

CONCLUSION : 2D video-based tracking could objectively evaluate gait in PD patients.

Shin Jung Hwan, Yu Ri, Ong Jed Noel, Lee Chan Young, Jeon Seung Ho, Park Hwanpil, Kim Han-Joon, Lee Jehee, Jeon Beomseok

2021-Apr-30

Parkinson disease, deep-learning, gait, video tracking

General General

A Novel Ensemble-based Classifier for Detecting the COVID-19 Disease for Infected Patients.

In Information systems frontiers : a journal of research and innovation

The recently discovered coronavirus, SARS-CoV-2, which was detected in Wuhan, China, has spread worldwide and is still being studied at the end of 2019. Detection of COVID-19 at an early stage is essential to provide adequate healthcare to affected patients and protect the uninfected community. This paper aims to design and develop a novel ensemble-based classifier to predict COVID-19 cases at a very early stage so that appropriate action can be taken by patients, doctors, health organizations, and the government. In this paper, a synthetic dataset of COVID-19 is generated by a dataset generation algorithm. A novel ensemble-based classifier of machine learning is employed on the COVID-19 dataset to predict the disease. A convex hull-based approach is also applied to the data to improve the proposed novel, ensemble-based classifier's accuracy and speed. The model is designed and developed through the python programming language and compares with the most popular classifier, i.e., Decision Tree, ID3, and support vector machine. The results indicate that the proposed novel classifier provides a more significant precision, kappa static, root means a square error, recall, F-measure, and accuracy.

Singh Prabh Deep, Kaur Rajbir, Singh Kiran Deep, Dhiman Gaurav

2021-Apr-25

Artificial intelligence, COVID-19, Corona virus, Ensemble classifier, Machine learning, Quality of service

General General

Federated Multi-View Learning for Private Medical Data Integration and Analysis

ArXiv Preprint

Along with the rapid expansion of information technology and digitalization of health data, there is an increasing concern on maintaining data privacy while garnering the benefits in medical field. Two critical challenges are identified: Firstly, medical data is naturally distributed across multiple local sites, making it difficult to collectively train machine learning models without data leakage. Secondly, in medical applications, data are often collected from different sources and views, resulting in heterogeneity and complexity that requires reconciliation. This paper aims to provide a generic Federated Multi-View Learning (FedMV) framework for multi-view data leakage prevention, which is based on different types of local data availability and enables to accommodate two types of problems: Vertical Federated Multi-View Learning (V-FedMV) and Horizontal Federated Multi-View Learning (H-FedMV). We experimented with real-world keyboard data collected from BiAffect study. The results demonstrated that the proposed FedMV approach can make full use of multi-view data in a privacy-preserving way, and both V-FedMV and H-FedMV methods perform better than their single-view and pairwise counterparts. Besides, the proposed model can be easily adapted to deal with multi-view sequential data in a federated environment, which has been modeled and experimentally studied. To the best of our knowledge, this framework is the first to consider both vertical and horizontal diversification in the multi-view setting, as well as their sequential federated learning.

Sicong Che, Hao Peng, Lichao Sun, Yong Chen, Lifang He

2021-05-04

General General

Convolutional neural networks for Alzheimer's disease detection on MRI images.

In Journal of medical imaging (Bellingham, Wash.)

Purpose: Detection of Alzheimer's disease (AD) on magnetic resonance imaging (MRI) using convolutional neural networks (CNNs), which is useful for detecting AD in its preliminary states. Approach: Our study implements and compares several deep models and configurations, including two-dimensional (2D) and three-dimensional (3D) CNNs and recurrent neural networks (RNNs). To use a 2D CNN on 3D MRI volumes, each MRI scan is split into 2D slices, neglecting the connection among 2D image slices in an MRI volume. Instead, a CNN model could be followed by an RNN in a way that the model of 2D CNN + RNN can understand the connection among sequences of 2D image slices for an MRI. The issue is that the feature extraction step in the 2D CNN is independent of classification in the RNN. To tackle this, 3D CNNs can be employed instead of 2D CNNs to make voxel-based decisions. Our study's main contribution is to introduce transfer learning from a dataset of 2D images to 3D CNNs. Results: The results on our MRI dataset indicate that sequence-based decisions improve the accuracy of slice-based decisions by 2% in classifying AD patients from healthy subjects. Also the 3D voxel-based method with transfer learning outperforms the other methods with 96.88% accuracy, 100% sensitivity, and 94.12% specificity. Conclusions: Several implementations and experiments using CNNs on MRI scans for AD detection demonstrated that the voxel-based method with transfer learning from ImageNet to MRI datasets using 3D CNNs considerably improved the results compared with the others.

Ebrahimi Amir, Luo Suhuai

2021-Mar

Alzheimer’s disease, convolutional neural network, deep learning, magnetic resonance imaging, transfer learning

Radiology Radiology

Algorithms for segmenting cerebral time-of-flight magnetic resonance angiograms from volunteers and anemic patients.

In Journal of medical imaging (Bellingham, Wash.)

Purpose: To evaluate six cerebral arterial segmentation algorithms in a set of patients with a wide range of hemodynamic characteristics to determine real-world performance. Approach: Time-of-flight magnetic resonance angiograms were acquired from 33 subjects: normal controls ( N = 11 ), sickle cell disease ( N = 11 ), and non-sickle anemia ( N = 11 ) using a 3 Tesla Philips Achieva scanner. Six segmentation algorithms were tested: (1) Otsu's method, (2) K-means, (3) region growing, (4) active contours, (5) minimum cost path, and (6) U-net machine learning. Segmentation algorithms were tested with two region-selection methods: global, which selects the entire volume; and local, which iteratively tracks the arteries. Five slices were manually segmented from each patient by two readers. Agreement between manual and automatic segmentation was measured using Matthew's correlation coefficient (MCC). Results: Median algorithm segmentation times ranged from 0.1 to 172.9 s for a single angiogram versus 10 h for manual segmentation. Algorithms had inferior performance to inter-observer vessel-based ( p < 0.0001 , MCC = 0.65 ) and voxel-based ( p < 0.0001 , MCC = 0.73 ) measurements. There were significant differences between algorithms ( p < 0.0001 ) and between patients ( p < 0.0042 ). Post-hoc analyses indicated (1) local minimum cost path performed best with vessel-based ( p = 0.0261 , MCC = 0.50 ) and voxel-based ( p = 0.0131 , MCC = 0.66 ) analyses; and (2) higher vessel-based performance in non-sickle anemia ( p = 0.0002 ) and lower voxel-based performance in sickle cell ( p = 0.0422 ) compared with normal controls. All reported MCCs are medians. Conclusions: The best-performing algorithm (local minimum cost path, voxel-based) had 9.59% worse performance than inter-observer agreement but was 3 orders of magnitude faster. Automatic segmentation was non-inferior in patients with sickle cell disease and superior in non-sickle anemia.

Saunders Alexander, King Kevin S, Blüml Stefan, Wood John C, Borzage Matthew

2021-Mar

anemia, image segmentation, time-of-flight magnetic resonance angiography

General General

Predicting Ovarian/Breast Cancer Pathogenic Risks of Human BRCA1 Gene Variants of Unknown Significance.

In BioMed research international ; h5-index 102.0

High-throughput sequencing is gaining popularity in clinical diagnoses, but more and more novel gene variants with unknown clinical significance are being found, giving difficulties to interpretations of people's genetic data, precise disease diagnoses, and the making of therapeutic strategies and decisions. In order to solve these issues, it is of critical importance to figure out ways to analyze and interpret such variants. In this work, BRCA1 gene variants with unknown clinical significance were identified from clinical sequencing data, and then, we developed machine learning models so as to predict the pathogenicity for variants with unknown clinical significance. Through performance benchmarking, we found that the optimized random forest model scored 0.85 in area under receiver operating characteristic curve, which outperformed other models. Finally, we applied the best random forest model to predict the pathogenicity of 6321 BRCA1 variants from both sequencing data and ClinVar database. As a result, we obtained the predictive pathogenic risks of BRCA1 variants of unknown significance.

Lin Hui-Heng, Xu Hongyan, Hu Hongbo, Ma Zhanzhong, Zhou Jie, Liang Qingyun

2021

General General

Classifying signals from a wearable accelerometer device to measure respiratory rate.

In ERJ open research

Background : Automatic measurement of respiratory rate in general hospital patients is difficult. Patient movement degrades the signal and variation of the breathing cycle means that accurate observation for ≥60 s is needed for adequate precision.

Methods : We studied acutely ill patients recently admitted to a teaching hospital. Breath duration was measured from a triaxial accelerometer attached to the chest wall and compared with a signal from a nasal cannula. We randomly divided the patient records into a training (n=54) and a test set (n=7). We used machine learning to train a neural network to select reliable signals, automatically identifying signal features associated with accurate measurement of respiratory rate. We used the test records to assess the accuracy of the device, indicated by the median absolute difference between respiratory rates, provided by the accelerometer and by the nasal cannula.

Results : In the test set of patients, machine classification of the respiratory signal reduced the median absolute difference (interquartile range) from 1.25 (0.56-2.18) to 0.48 (0.30-0.78) breaths per min. 50% of the recording periods were rejected as unreliable and in one patient, only 10% of the signal time was classified as reliable. However, even only 10% of observation time would allow accurate measurement for 6 min in an hour of recording, giving greater reliability than nurse charting, which is based on much less observation time.

Conclusion : Signals from a body-mounted accelerometer yield accurate measures of respiratory rate, which could improve automatic illness scoring in adult hospital patients.

Drummond Gordon B, Fischer Darius, Lees Margaret, Bates Andrew, Mann Janek, Arvind D K

2021-Apr

Cardiology Cardiology

The Incremental Prognostic Value of the Clinical Residual SYNTAX Score for Patients With Chronic Renal Insufficiency Undergoing Percutaneous Coronary Intervention.

In Frontiers in cardiovascular medicine

Background: The residual SYNTAX score (RSS) is considered a powerful prognostic indicator for determining a reasonable revascularization strategy in patients undergoing percutaneous coronary intervention (PCI), but the absence of clinical parameters is one of the limitations of RSS, especially in the chronic renal insufficiency (CRI) comorbidity setting. The present work aimed to investigate the incremental prognostic value of clinical residual SYNTAX score (CRSS) compared with RSS in CRI cases after PCI. Methods: Totally 2,468 consecutive CRI cases who underwent PCI from January 2014 to September 2017 were included in this retrospective analysis. CRSS was obtained by multiplying RSS by the modified ACEF score. Individuals with CRSS >0 were considered to have incomplete revascularization and stratified by CRSS tertiles, the remaining cases constituted the complete revascularization (CR) group. The outcomes between these groups were compared. Results: At a median follow-up of 3 years, compared with CR group, individuals with CRSS >12 showed elevated rates of all clinical outcomes, and those with CRSS ≤ 12 showed similar all-cause and cardiac mortality rates. In multivariable analysis, CRSS was a powerful independent predictive factor of all clinical outcomes. The net reclassification improvement levels of CRSS over RSS for all-cause and cardiac mortality rates were 10.3% (p = 0.007) and 16.4% (p < 0.001), respectively. Compared with RSS, CRSS markedly ameliorated all-cause and cardiac mortality risk stratification. Conclusions: Compared with RSS, CRSS has incremental predictability for long-term all-cause and cardiac mortality in CRI cases following PCI.

Yan Liqiu, Li Peiyao, Wang Yabin, Han Dong, Li Sulei, Jiang Min, Cao Xufen, Cao Feng

2021

chronic renal insufficiency, clinical residual SYNTAX score, coronary artery disease, percutaneous coronary intervention, residual SYNTAX score

General General

Low-Density Lipoprotein Cholesterol 4: The Notable Risk Factor of Coronary Artery Disease Development.

In Frontiers in cardiovascular medicine

Background: Coronary artery disease (CAD) is the leading cause of death worldwide, which has a long asymptomatic period of atherosclerosis. Thus, it is crucial to develop efficient strategies or biomarkers to assess the risk of CAD in asymptomatic individuals. Methods: A total of 356 consecutive CAD patients and 164 non-CAD controls diagnosed using coronary angiography were recruited. Blood lipids, other baseline characteristics, and clinical information were investigated in this study. In addition, low-density lipoprotein cholesterol (LDL-C) subfractions were classified and quantified using the Lipoprint system. Based on these data, we performed comprehensive analyses to investigate the risk factors for CAD development and to predict CAD risk. Results: Triglyceride, LDLC-3, LDLC-4, LDLC-5, LDLC-6, and total small and dense LDL-C were significantly higher in the CAD patients than those in the controls, whereas LDLC-1 and high-density lipoprotein cholesterol (HDL-C) had significantly lower levels in the CAD patients. Logistic regression analysis identified male [odds ratio (OR) = 2.875, P < 0.001], older age (OR = 1.018, P = 0.025), BMI (OR = 1.157, P < 0.001), smoking (OR = 4.554, P < 0.001), drinking (OR = 2.128, P < 0.016), hypertension (OR = 4.453, P < 0.001), and diabetes mellitus (OR = 8.776, P < 0.001) as clinical risk factors for CAD development. Among blood lipids, LDLC-3 (OR = 1.565, P < 0.001), LDLC-4 (OR = 3.566, P < 0.001), and LDLC-5 (OR = 6.866, P < 0.001) were identified as risk factors. To predict CAD risk, six machine learning models were constructed. The XGboost model showed the highest AUC score (0.945121), which could distinguish CAD patients from the controls with a high accuracy. LDLC-4 played the most important role in model construction. Conclusions: The established models showed good performance for CAD risk prediction, which can help screen high-risk CAD patients in asymptomatic population, so that further examination and prevention treatment might be taken before any sudden or serious event.

Wu Dongmei, Yang Qiuju, Su Baohua, Hao Jia, Ma Huirong, Yuan Weilan, Gao Junhui, Ding Feifei, Xu Yue, Wang Huifeng, Zhao Jiangman, Li Bingqiang

2021

LDL-C subfractions, coronary angiography, coronary artery disease, machine learning, risk factors

General General

Perspective: Wearable Internet of Medical Things for Remote Tracking of Symptoms, Prediction of Health Anomalies, Implementation of Preventative Measures, and Control of Virus Spread During the Era of COVID-19.

In Frontiers in robotics and AI

The COVID-19 pandemic has highly impacted the communities globally by reprioritizing the means through which various societal sectors operate. Among these sectors, healthcare providers and medical workers have been impacted prominently due to the massive increase in demand for medical services under unprecedented circumstances. Hence, any tool that can help the compliance with social guidelines for COVID-19 spread prevention will have a positive impact on managing and controlling the virus outbreak and reducing the excessive burden on the healthcare system. This perspective article disseminates the perspectives of the authors regarding the use of novel biosensors and intelligent algorithms embodied in wearable IoMT frameworks for tackling this issue. We discuss how with the use of smart IoMT wearables certain biomarkers can be tracked for detection of COVID-19 in exposed individuals. We enumerate several machine learning algorithms which can be used to process a wide range of collected biomarkers for detecting (a) multiple symptoms of SARS-CoV-2 infection and (b) the dynamical likelihood of contracting the virus through interpersonal interaction. Eventually, we enunciate how a systematic use of smart wearable IoMT devices in various social sectors can intelligently help controlling the spread of COVID-19 in communities as they enter the reopening phase. We explain how this framework can benefit individuals and their medical correspondents by introducing Systems for Symptom Decoding (SSD), and how the use of this technology can be generalized on a societal level for the control of spread by introducing Systems for Spread Tracing (SST).

Mehrdad Sarmad, Wang Yao, Atashzar S Farokh

2021

AI for health, COVID-19, IoMT, smart connected health, smart wearables, spread control, symptom tracking, telemedicine

General General

Detecting Protein Communities in Native Cell Extracts by Machine Learning: A Structural Biologist's Perspective.

In Frontiers in molecular biosciences

Native cell extracts hold great promise for understanding the molecular structure of ordered biological systems at high resolution. This is because higher-order biomolecular interactions, dubbed as protein communities, may be retained in their (near-)native state, in contrast to extensively purifying or artificially overexpressing the proteins of interest. The distinct machine-learning approaches are applied to discover protein-protein interactions within cell extracts, reconstruct dedicated biological networks, and report on protein community members from various organisms. Their validation is also important, e.g., by the cross-linking mass spectrometry or cell biology methods. In addition, the cell extracts are amenable to structural analysis by cryo-electron microscopy (cryo-EM), but due to their inherent complexity, sorting structural signatures of protein communities derived by cryo-EM comprises a formidable task. The application of image-processing workflows inspired by machine-learning techniques would provide improvements in distinguishing structural signatures, correlating proteomic and network data to structural signatures and subsequently reconstructed cryo-EM maps, and, ultimately, characterizing unidentified protein communities at high resolution. In this review article, we summarize recent literature in detecting protein communities from native cell extracts and identify the remaining challenges and opportunities. We argue that the progress in, and the integration of, machine learning, cryo-EM, and complementary structural proteomics approaches would provide the basis for a multi-scale molecular description of protein communities within native cell extracts.

Kyrilis Fotis L, Belapure Jaydeep, Kastritis Panagiotis L

2021

cellular homogenates, convolutional neural network, cryo-EM, mass spectrometry, metabolons, protein–protein interactions, random forest, structural biology

Radiology Radiology

Nanogenomics and Artificial Intelligence: A Dynamic Duo for the Fight Against Breast Cancer.

In Frontiers in molecular biosciences

Application software is utilized to aid in the diagnosis of breast cancer. Yet, recent advances in artificial intelligence (AI) are addressing challenges related to the detection, classification, and monitoring of different types of tumors. AI can apply deep learning algorithms to perform automated analysis on mammographic or histologic examinations. Large volume of data generated by digitalized mammogram or whole-slide images can be interoperated through advanced machine learning. This enables fast evaluation of every tissue patch on an image, resulting in a quicker more sensitivity, and more reproducible diagnoses compared to human performance. On the other hand, cancer cell-exosomes which are extracellular vesicles released by cancer cells into the blood circulation, are being explored as cancer biomarker. Recent studies on cancer-exosome-content revealed that the encapsulated miRNA and other biomolecules are indicative of tumor sub-type, possible metastasis and prognosis. Thus, theoretically, through nanogenomicas, a profile of each breast tumor sub-type, estrogen receptor status, and potential metastasis site can be constructed. Then, a laboratory instrument, fitted with an AI program, can be used to diagnose suspected patients by matching their sera miRNA and biomolecules composition with the available template profiles. In this paper, we discuss the advantages of establishing a nanogenomics-AI-based breast cancer diagnostic approach, compared to the gold standard radiology or histology based approaches that are currently being adapted to AI. Also, we discuss the advantages of building the diagnostic and prognostic biomolecular profiles for breast cancers based on the exosome encapsulated content, rather than the free circulating miRNA and other biomolecules.

Al-Sowayan Batla S, Al-Shareeda Alaa T

2021

artificial intelligence, cancer, diagnosis, exosome, machine learning, screening

Dermatology Dermatology

Diagnosis of Onychomycosis: From Conventional Techniques and Dermoscopy to Artificial Intelligence.

In Frontiers in medicine

Onychomycosis is a common fungal nail infection. Accurate diagnosis is critical as onychomycosis is transmissible between humans and impacts patients' quality of life. Combining clinical examination with mycological testing ensures accurate diagnosis. Conventional diagnostic techniques, including potassium hydroxide testing, fungal culture and histopathology of nail clippings, detect fungal species within nails. New diagnostic tools have been developed recently which either improve detection of onychomycosis clinically, including dermoscopy, reflectance confocal microscopy and artificial intelligence, or mycologically, such as molecular assays. Dermoscopy is cost-effective and non-invasive, allowing clinicians to discern microscopic features of onychomycosis and fungal melanonychia. Reflectance confocal microscopy enables clinicians to observe bright filamentous septate hyphae at near histologic resolution by the bedside. Artificial intelligence may prompt patients to seek further assessment for nails that are suspicious for onychomycosis. This review evaluates the current landscape of diagnostic techniques for onychomycosis.

Lim Sophie Soyeon, Ohn Jungyoon, Mun Je-Ho

2021

artificial intelligence, dermoscopy, diagnosis, diagnostic imaging, fungi, onychomycosis, pathology, reflectance confocal microscopy

Surgery Surgery

The Year in Perioperative Echocardiography: Selected Highlights from 2020.

In Journal of cardiothoracic and vascular anesthesia ; h5-index 35.0

This article is the fifth of an annual series reviewing the research highlights of the year pertaining to the subspecialty of perioperative echocardiography for the Journal of Cardiothoracic and Vascular Anesthesia. The authors thank Editor-in-Chief Dr. Kaplan and the editorial board for the opportunity to continue this series. In most cases, these will be research articles that are targeted at the perioperative echocardiography diagnosis and treatment of patients after cardiothoracic surgery; but in some cases, these articles will target the use of perioperative echocardiography in general.

Khoche Swapnil, Hashmi Nazish, Bronshteyn Yuriy S, Choi Christine, Poorsattar Sophia, Maus Timothy M

2021-Mar-27

artificial intelligence, focused cardiac ultrasound, perioperative echocardiography, point-of-care ultrasound, surgical decision-making, transesophageal echocardiography safety

General General

Optimization of Thermal and Structural Design in Lithium-Ion Batteries to Obtain Energy Efficient Battery Thermal Management System (BTMS): A Critical Review.

In Archives of computational methods in engineering : state of the art reviews

Covid-19 has given one positive perspective to look at our planet earth in terms of reducing the air and noise pollution thus improving the environmental conditions globally. This positive outcome of pandemic has given the indication that the future of energy belong to green energy and one of the emerging source of green energy is Lithium-ion batteries (LIBs). LIBs are the backbone of the electric vehicles but there are some major issues faced by the them like poor thermal performance, thermal runaway, fire hazards and faster rate of discharge under low and high temperature environment,. Therefore to overcome these problems most of the researchers have come up with new methods of controlling and maintaining the overall thermal performance of the LIBs. The present review paper mainly is focused on optimization of thermal and structural design parameters of the LIBs under different BTMSs. The optimized BTMS generally demonstrated in this paper are maximum temperature of battery cell, battery pack or battery module, temperature uniformity, maximum or average temperature difference, inlet temperature of coolant, flow velocity, and pressure drop. Whereas the major structural design optimization parameters highlighted in this paper are type of flow channel, number of channels, length of channel, diameter of channel, cell to cell spacing, inlet and outlet plenum angle and arrangement of channels. These optimized parameters investigated under different BTMS heads such as air, PCM (phase change material), mini-channel, heat pipe, and water cooling are reported profoundly in this review article. The data are categorized and the results of the recent studies are summarized for each method. Critical review on use of various optimization algorithms (like ant colony, genetic, particle swarm, response surface, NSGA-II, etc.) for design parameter optimization are presented and categorized for different BTMS to boost their objectives. The single objective optimization techniques helps in obtaining the optimal value of important design parameters related to the thermal performance of battery cooling systems. Finally, multi-objective optimization technique is also discussed to get an idea of how to get the trade-off between the various conflicting parameters of interest such as energy, cost, pressure drop, size, arrangement, etc. which is related to minimization and thermal efficiency/performance of the battery system related to maximization. This review will be very helpful for researchers working with an objective of improving the thermal performance and life span of the LIBs.

Fayaz H, Afzal Asif, Samee A D Mohammed, Soudagar Manzoore Elahi M, Akram Naveed, Mujtaba M A, Jilte R D, Islam Md Tariqul, Ağbulut Ümit, Saleel C Ahamed

2021-Apr-26

General General

Conversational Machine Reading Comprehension for Vietnamese Healthcare Texts

ArXiv Preprint

Machine reading comprehension (MRC) is a sub-field in natural language processing or computational linguistics. MRC aims to help computers understand unstructured texts and then answer questions related to them. In this paper, we present a new Vietnamese corpus for conversational machine reading comprehension (ViCoQA), consisting of 10,000 questions with answers over 2,000 conversations about health news articles. We analyze ViCoQA in depth with different linguistic aspects. Then, we evaluate several baseline models about dialogue and reading comprehension on the ViCoQA corpus. The best model obtains an F1 score of 45.27%, which is 30.91 points behind human performance (76.18%), indicating that there is ample room for improvement.

Son T. Luu, Mao Nguyen Bui, Loi Duc Nguyen, Khiem Vinh Tran, Kiet Van Nguyen, Ngan Luu-Thuy Nguyen

2021-05-04

Dermatology Dermatology

A Deep Learning Based Framework for Diagnosing Multiple Skin Diseases in a Clinical Environment.

In Frontiers in medicine

Background: Numerous studies have attempted to apply artificial intelligence (AI) in the dermatological field, mainly on the classification and segmentation of various dermatoses. However, researches under real clinical settings are scarce. Objectives: This study was aimed to construct a novel framework based on deep learning trained by a dataset that represented the real clinical environment in a tertiary class hospital in China, for better adaptation of the AI application in clinical practice among Asian patients. Methods: Our dataset was composed of 13,603 dermatologist-labeled dermoscopic images, containing 14 categories of diseases, namely lichen planus (LP), rosacea (Rosa), viral warts (VW), acne vulgaris (AV), keloid and hypertrophic scar (KAHS), eczema and dermatitis (EAD), dermatofibroma (DF), seborrheic dermatitis (SD), seborrheic keratosis (SK), melanocytic nevus (MN), hemangioma (Hem), psoriasis (Pso), port wine stain (PWS), and basal cell carcinoma (BCC). In this study, we applied Google's EfficientNet-b4 with pre-trained weights on ImageNet as the backbone of our CNN architecture. The final fully-connected classification layer was replaced with 14 output neurons. We added seven auxiliary classifiers to each of the intermediate layer groups. The modified model was retrained with our dataset and implemented using Pytorch. We constructed saliency maps to visualize our network's attention area of input images for its prediction. To explore the visual characteristics of different clinical classes, we also examined the internal image features learned by the proposed framework using t-SNE (t-distributed Stochastic Neighbor Embedding). Results: Test results showed that the proposed framework achieved a high level of classification performance with an overall accuracy of 0.948, a sensitivity of 0.934 and a specificity of 0.950. We also compared the performance of our algorithm with three most widely used CNN models which showed our model outperformed existing models with the highest area under curve (AUC) of 0.985. We further compared this model with 280 board-certificated dermatologists, and results showed a comparable performance level in an 8-class diagnostic task. Conclusions: The proposed framework retrained by the dataset that represented the real clinical environment in our department could accurately classify most common dermatoses that we encountered during outpatient practice including infectious and inflammatory dermatoses, benign and malignant cutaneous tumors.

Zhu Chen-Yu, Wang Yu-Kun, Chen Hai-Peng, Gao Kun-Lun, Shu Chang, Wang Jun-Cheng, Yan Li-Feng, Yang Yi-Guang, Xie Feng-Ying, Liu Jie

2021

artificial intelligence, convolutional neural networks, deep learning, dermatology, dermoscopy, skin diseases, skin imaging

Surgery Surgery

Hyperglycemia-Induced Dysregulated Fusion Intermediates in Insulin-Secreting Cells Visualized by Super-Resolution Microscopy.

In Frontiers in cell and developmental biology

Impaired insulin release is a hallmark of type 2 diabetes and is closely related to chronically elevated glucose concentrations, known as "glucotoxicity." However, the molecular mechanisms by which glucotoxicity impairs insulin secretion remain poorly understood. In addition to known kiss-and-run and kiss-and-stay fusion events in INS-1 cells, ultrafast Hessian structured illumination microscopy (Hessian SIM) enables full fusion to be categorized according to the newly identified structures, such as ring fusion (those with enlarged pores) or dot fusion (those without apparent pores). In addition, we identified four fusion intermediates during insulin exocytosis: initial pore opening, vesicle collapse, enlarged pore formation, and final pore dilation. Long-term incubation in supraphysiological doses of glucose reduced exocytosis in general and increased the occurrence of kiss-and-run events at the expense of reduced full fusion. In addition, hyperglycemia delayed pore opening, vesicle collapse, and enlarged pore formation in full fusion events. It also reduced the size of apparently enlarged pores, all of which contributed to the compromised insulin secretion. These phenotypes were mostly due to the hyperglycemia-induced reduction in syntaxin-1A (Stx-1A) and SNAP-25 protein, since they could be recapitulated by the knockdown of endogenous Stx-1A and SNAP-25. These findings suggest essential roles for the vesicle fusion type and intermediates in regulating insulin secretion from pancreatic beta cells in normal and disease conditions.

Yang Guoyi, Li Liuju, Liu Yanmei, Liang Kuo, Wei Lisi, Chen Liangyi

2021

SNARE, exocytosis, fusion pore, glucotoxicity, insulin, secretory vesicle, structured illumination microscopy