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

Cardiology Cardiology

Elicitation of domain knowledge for a machine learning model for paediatric critical illness in South Africa.

In Frontiers in pediatrics

OBJECTIVES : Delays in identification, resuscitation and referral have been identified as a preventable cause of avoidable severity of illness and mortality in South African children. To address this problem, a machine learning model to predict a compound outcome of death prior to discharge from hospital and/or admission to the PICU was developed. A key aspect of developing machine learning models is the integration of human knowledge in their development. The objective of this study is to describe how this domain knowledge was elicited, including the use of a documented literature search and Delphi procedure.

DESIGN : A prospective mixed methodology development study was conducted that included qualitative aspects in the elicitation of domain knowledge, together with descriptive and analytical quantitative and machine learning methodologies.

SETTING : A single centre tertiary hospital providing acute paediatric services.

PARTICIPANTS : Three paediatric intensivists, six specialist paediatricians and three specialist anaesthesiologists.

INTERVENTIONS : None.

MEASUREMENTS AND MAIN RESULTS : The literature search identified 154 full-text articles reporting risk factors for mortality in hospitalised children. These factors were most commonly features of specific organ dysfunction. 89 of these publications studied children in lower- and middle-income countries. The Delphi procedure included 12 expert participants and was conducted over 3 rounds. Respondents identified a need to achieve a compromise between model performance, comprehensiveness and veracity and practicality of use. Participants achieved consensus on a range of clinical features associated with severe illness in children. No special investigations were considered for inclusion in the model except point-of-care capillary blood glucose testing. The results were integrated by the researcher and a final list of features was compiled.

CONCLUSION : The elicitation of domain knowledge is important in effective machine learning applications. The documentation of this process enhances rigour in such models and should be reported in publications. A documented literature search, Delphi procedure and the integration of the domain knowledge of the researchers contributed to problem specification and selection of features prior to feature engineering, pre-processing and model development.

Pienaar Michael A, Sempa Joseph B, Luwes Nicolaas, George Elizabeth C, Brown Stephen C

2023

children, critical care, domain knowledge, machine learning, severity of illness, triage

General General

Probing the relevance of the accelerated aging mouse line SAMP8 as a model for certain types of neuropsychiatric symptoms in dementia.

In Frontiers in psychiatry

INTRODUCTION : People with dementia (PwD) often present with neuropsychiatric symptoms (NPS). NPS are of substantial burden to the patients, and current treatment options are unsatisfactory. Investigators searching for novel medications need animal models that present disease-relevant phenotypes and can be used for drug screening. The Senescence Accelerated Mouse-Prone 8 (SAMP8) strain shows an accelerated aging phenotype associated with neurodegeneration and cognitive decline. Its behavioural phenotype in relation to NPS has not yet been thoroughly investigated. Physical and verbal aggression in reaction to the external environment (e.g., interaction with the caregiver) is one of the most prevalent and debilitating NPS occurring in PwD. Reactive aggression can be studied in male mice using the Resident-Intruder (R-I) test. SAMP8 mice are known to be more aggressive than the Senescence Accelerated Mouse-Resistant 1 (SAMR1) control strain at specific ages, but the development of the aggressive phenotype over time, is still unknown.

METHODS : In our study, we performed a longitudinal, within-subject, assessment of aggressive behaviour of male SAMP8 and SAMR1 mice at 4, 5, 6 and 7 months of age. Aggressive behaviour from video recordings of the R-I sessions was analysed using an in-house developed behaviour recognition software.

RESULTS : SAMP8 mice were more aggressive relative to SAMR1 mice starting at 5 months of age, and the phenotype was still present at 7 months of age. Treatment with risperidone (an antipsychotic frequently used to treat agitation in clinical practice) reduced aggression in both strains. In a three-chamber social interaction test, SAMP8 mice also interacted more fervently with male mice than SAMR1, possibly because of their aggression-seeking phenotype. They did not show any social withdrawal.

DISCUSSION : Our data support the notion that SAMP8 mice might be a useful preclinical tool to identify novel treatment options for CNS disorders associated with raised levels of reactive aggression such as dementia.

Bergamini Giorgio, Massinet Helene, Hart Aaron, Durkin Sean, Pierlot Gabin, Steiner Michel Alexander

2023

SAMP8 mice, aggression, aging, agitation, deep learning, dementia, sociability

Radiology Radiology

Radiomics model using preoperative computed tomography angiography images to differentiate new from old emboli of acute lower limb arterial embolism.

In Open medicine (Warsaw, Poland)

Our purpose was to devise a radiomics model using preoperative computed tomography angiography (CTA) images to differentiate new from old emboli of acute lower limb arterial embolism. 57 patients (95 regions of interest; training set: n = 57; internal validation set: n = 38) with femoral popliteal acute lower limb arterial embolism confirmed by pathology and with preoperative CTA images were retrospectively analyzed. We selected the best prediction model according to the model performance tested by area under the curve (AUC) analysis across 1,000 iterations of prediction from three most common machine learning methods: support vector machine, feed-forward neural network (FNN), and random forest, through several steps of feature selection. Then, the selected best model was also validated in an external validation dataset (n = 24). The established radiomics signature had good predictive efficacy. FNN exhibited the best model performance on the training and validation groups: its AUC value was 0.960 (95% CI, 0.899-1). The accuracy of this model was 89.5%, and its sensitivity and specificity were 0.938 and 0.864, respectively. The AUC of external validation dataset was 0.793. Our radiomics model based on preoperative CTA images is valuable. The radiomics approach of preoperative CTA to differentiate new emboli from old is feasible.

Liu Rong, Yang Junlin, Zhang Wei, Li Xiaobo, Shi Dai, Cai Wu, Zhang Yue, Fan Guohua, Li Chenglong, Jiang Zhen

2023

X-ray computed tomography, arterial embolization, radiomics, thrombosis

Radiology Radiology

CoviExpert: COVID-19 detection from chest X-ray using CNN.

In Measurement. Sensors

COVID-19 continues to threaten the world with its impact and severity. This pandemic has created a sense of havoc and shook the world stretching the medical fraternity to an unimaginable extent, who are now facing fatigue and exhaustion. Due to the rapid increase in cases all across the globe demanding extensive medical care, people are hunting for resources like testing facilities, medical drugs and even hospital beds. Even people with mild to moderate infection are panicking and mentally giving up due to anxiety and desperation. To combat these issues, it is necessary to find an inexpensive and faster way to save lives and bring about a much-needed change. The most fundamental way through which this can be achieved is with the help of radiology which involves examination of Chest X rays. They are primarily used for the diagnosis of this disease. But due to panic and severity of this disease a recent trend of performing CT scans has been observed. This has been under scrutiny since it exposes patients to a very high level of radiation known to increase the probability of cancer. As quoted by the AIIMS Director, one CT scan is equivalent to around 300-400 Chest X-rays. Also, it is relatively a much costlier testing method. Hence, in this report, we have presented a Deep learning approach which can detect covid 19 positive cases from Chest X ray images. It involves creation of a Deep learning based Convolutional Neural Network (CNN) using Keras (python library) and integrating the model with a front-end user interface for ease of use. This leads up to the creation of a software which we have named as CoviExpert. It uses the sequential Keras model which is built layer by layer. All the layers are trained independently to make independent predictions which are then combined to give the final output. 1584 images of Chest X-rays of both COVID-19 positive and negative patients have been used as training data. 177 images have been used as testing data. The proposed approach gives a classification accuracy of 99%. CoviExpert can be used on any device by any medical professional to detect Covid positive patients within a few seconds.

Arivoli A, Golwala Devdatt, Reddy Rayirth

2022-Oct

CNN, COVID-19, CT scan, CoviExpert, Deep learning, X-ray

General General

Synthetic computed tomography generation for abdominal adaptive radiotherapy using low-field magnetic resonance imaging.

In Physics and imaging in radiation oncology

BACKGROUND AND PURPOSE : Magnetic Resonance guided Radiotherapy (MRgRT) still needs the acquisition of Computed Tomography (CT) images and co-registration between CT and Magnetic Resonance Imaging (MRI). The generation of synthetic CT (sCT) images from the MR data can overcome this limitation. In this study we aim to propose a Deep Learning (DL) based approach for sCT image generation for abdominal Radiotherapy using low field MR images.

MATERIALS AND METHODS : CT and MR images were collected from 76 patients treated on abdominal sites. U-Net and conditional Generative Adversarial Network (cGAN) architectures were used to generate sCT images. Additionally, sCT images composed of only six bulk densities were generated with the aim of having a Simplified sCT.Radiotherapy plans calculated using the generated images were compared to the original plan in terms of gamma pass rate and Dose Volume Histogram (DVH) parameters.

RESULTS : sCT images were generated in 2 s and 2.5 s with U-Net and cGAN architectures respectively.Gamma pass rates for 2%/2mm and 3%/3mm criteria were 91% and 95% respectively. Dose differences within 1% for DVH parameters on the target volume and organs at risk were obtained.

CONCLUSION : U-Net and cGAN architectures are able to generate abdominal sCT images fast and accurately from low field MRI.

Garcia Hernandez Armando, Fau Pierre, Wojak Julien, Mailleux Hugues, Benkreira Mohamed, Rapacchi Stanislas, Adel Mouloud

2023-Jan

Deep Learning, Low-field MRI, MR-only treatment planning, Synthetic CT

Internal Medicine Internal Medicine

Screening the risk of obstructive sleep apnea by utilizing supervised learning techniques based on anthropometric features and snoring events.

In Digital health

OBJECTIVES : Obstructive sleep apnea (OSA) is typically diagnosed by polysomnography (PSG). However, PSG is time-consuming and has some clinical limitations. This study thus aimed to establish machine learning models to screen for the risk of having moderate-to-severe and severe OSA based on easily acquired features.

METHODS : We collected PSG data on 3529 patients from Taiwan and further derived the number of snoring events. Their baseline characteristics and anthropometric measures were obtained, and correlations among the collected variables were investigated. Next, six common supervised machine learning techniques were utilized, including random forest (RF), extreme gradient boosting (XGBoost), k-nearest neighbor (kNN), support vector machine (SVM), logistic regression (LR), and naïve Bayes (NB). First, data were independently separated into a training and validation dataset (80%) and a test dataset (20%). The approach with the highest accuracy in the training and validation phase was employed to classify the test dataset. Next, feature importance was investigated by calculating the Shapley value of every factor, which represented the impact on OSA risk screening.

RESULTS : The RF produced the highest accuracy (of >70%) in the training and validation phase in screening for both OSA severities. Hence, we employed the RF to classify the test dataset, and results showed a 79.32% accuracy for moderate-to-severe OSA and 74.37% accuracy for severe OSA. Snoring events and the visceral fat level were the most and second most essential features of screening for OSA risk.

CONCLUSIONS : The established model can be considered for screening for the risk of having moderate-to-severe or severe OSA.

Tsai Cheng-Yu, Liu Wen-Te, Hsu Wen-Hua, Majumdar Arnab, Stettler Marc, Lee Kang-Yun, Cheng Wun-Hao, Wu Dean, Lee Hsin-Chien, Kuan Yi-Chun, Wu Cheng-Jung, Lin Yi-Chih, Ho Shu-Chuan

2023

Obstructive sleep apnea, Shapley value, anthropometric measure, machine learning, snoring event