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

oncology Oncology

Clinical validation of a machine-learning derived signature predictive of outcomes from first-line oxaliplatin-based chemotherapy in advanced colorectal cancer.

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

PURPOSE : FOLFOX, FOLFIRI, or FOLFOXIRI chemotherapy with bevacizumab (BV) are considered standard 1st line treatment options for patients with metastatic colorectal cancer (mCRC). We developed and validated a molecular signature predictive of efficacy of oxaliplatin-based chemotherapy combined with BV in patients with mCRC.

EXPERIMENTAL DESIGN : A machine-learning approach was applied and tested on clinical and NGS data from a real-world evidence (RWE) data set and samples from the prospective TRIBE2 study resulting in identification of a molecular signature - FOLFOXai Algorithm training considered time-to-next-treatment (TTNT). Validation studies used TTNT, PFS and overall survival (OS) as the primary endpoints.

RESULTS : A 67 gene signature was cross-validated in a training cohort (N=105) which demonstrated the ability of FOLFOXai to distinguish FOLFOX-treated mCRC patients with increased benefit (IB) from those with decreased benefit (DB). The signature was predictive of TTNT and OS in an independent RWE dataset of 412 patients who had received FOLFOX/BV in 1st line and inversely predictive of survival in RWE data from 55 patients who had received 1st line FOLFIRI. Blinded analysis of TRIBE2 samples confirmed that FOLFOXai was predictive of OS in both oxaliplatin-containing arms (FOLFOX HR=0.629, p=0.04 and FOLFOXIRI HR=0.483, p=0.02). FOLFOXai was also predictive of treatment benefit from oxaliplatin-containing regimens in advanced esophageal/gastro-esophageal junction cancers (EC/GEJC) as well as pancreatic ductal adenocarcinoma (PDAC).

CONCLUSIONS : Application of FOLFOXai could lead to improvements of treatment outcomes for patients with mCRC and other cancers since patients predicted to have less benefit from oxaliplatin-containing regimens might benefit from alternative regimens.

Abraham Jim P, Magee Daniel, Cremolini Chiara, Antoniotti Carlotta, Halbert David D, Xiu Joanne, Stafford Phillip, Berry Donald A, Oberley Matthew J, Shields Anthony F, Marshall John L, Salem Mohamed E, Falcone Alfredo, Grothey Axel, Hall Michael J, Venook Alan P, Lenz Heinz-Josef, Helmstetter Anthony, Korn W Michael, Spetzler David B

2020-Dec-08

General General

Identifying Patients at Risk for Diaphragm Atrophy During Mechanical Ventilation Using Routinely Available Clinical Data.

In Respiratory care ; h5-index 37.0

BACKGROUND : Diaphragmatic respiratory effort during mechanical ventilation is an important determinant of patient outcome, but direct measurement of diaphragmatic contractility requires specialized instrumentation and technical expertise. We sought to determine whether routinely collected clinical variables can predict diaphragmatic contractility and stratify the risk of diaphragm atrophy.

METHODS : We conducted a secondary analysis of a prospective cohort study on diaphragm ultrasound in mechanically ventilated subjects. Clinical variables, such as breathing frequency, ventilator settings, and blood gases, were recorded longitudinally. Machine learning techniques were used to identify variables predicting diaphragm contractility and stratifying the risk of diaphragm atrophy (> 10% decrease in thickness from baseline). Performance of the variables was evaluated in mixed-effects logistic regression and random-effects tree models using the area under the receiver operating characteristic curve.

RESULTS : Measurements were available for 761 study days in 191 subjects, of whom 73 (38%) developed diaphragm atrophy. No routinely collected clinical variable, alone or in combination, could accurately predict either diaphragm contractility or the development of diaphragm atrophy (model area under the receiver operating characteristic curve 0.63-0.75). The risk of diaphragm atrophy was not significantly different according to the presence or absence of patient-triggered breaths (38.3% vs 38.6%; odds ratio 1.01, 95% CI 0.05-2.03). Diaphragm thickening fraction < 15% during either of the first 2 d of the study was associated with a higher risk of atrophy (44.6% vs 26.1%; odds ratio 2.28, 95% CI 1.05-4.95).

CONCLUSIONS : Diaphragmatic contractility and the risk of diaphragm atrophy could not be reliably determined from routinely collected clinical variables and ventilator settings. A single measurement of diaphragm thickening fraction measured within 48 h of initiating mechanical ventilation can be used to stratify the risk of diaphragm atrophy during mechanical ventilation.

Urner Martin, Mitsakakis Nicholas, Vorona Stefannie, Chen Lu, Sklar Michael C, Dres Martin, Rubenfeld Gordon D, Brochard Laurent J, Ferguson Niall D, Fan Eddy, Goligher Ewan C

2020-Dec-08

diaphragm atrophy, diaphragm thickening, diaphragm thickening fraction, machine learning, random forest, spontaneous breathing

General General

Neuronal differentiation strategies: insights from single-cell sequencing and machine learning.

In Development (Cambridge, England)

Neuronal replacement therapies rely on the in vitro differentiation of specific cell types from embryonic or induced pluripotent stem cells, or on the direct reprogramming of differentiated adult cells via the expression of transcription factors or signaling molecules. The factors used to induce differentiation or reprogramming are often identified by informed guesses based on differential gene expression or known roles for these factors during development. Moreover, differentiation protocols usually result in partly differentiated cells or the production of a mix of cell types. In this Hypothesis article, we suggest that, to overcome these inefficiencies and improve neuronal differentiation protocols, we need to take into account the developmental history of the desired cell types. Specifically, we present a strategy that uses single-cell sequencing techniques combined with machine learning as a principled method to select a sequence of programming factors that are important not only in adult neurons but also during differentiation.

Konstantinides Nikolaos, Desplan Claude

2020-Dec-08

In vitro differentiation, Machine learning, Neuronal development, Neuronal differentiation protocols, Neuronal replacement therapy, Single-cell sequencing

Cardiology Cardiology

An 8-layer residual U-Net with deep supervision for segmentation of the left ventricle in cardiac CT angiography.

In Computer methods and programs in biomedicine

BACKGROUND AND OBJECTIVES : Accurate segmentation of left ventricle (LV) is a fundamental step in evaluation of cardiac function. Cardiac CT angiography (CCTA) has become an important clinical diagnostic method for cardio-vascular disease (CVD) due to its non-invasive, short exam time, and low cost. To obtain the segmentation of the LV in CCTA scans, we present a deep learning method based on an 8-layer residual U-Net with deep supervision.

METHODS : Based on the original 4-layer U-Net, our method deepened the network to eight layers, which increased the fitting capacity of the network, thus greatly improved its LV recognition capability. Residual blocks were incorporated to optimize the network from the increased depth. Auxiliary paths as deep supervision were introduced to supervise the intermediate information to improve the segmentation quality. In this study, we collected CCTA scans of 100 patients. Eighty patients with 1600 discrete slices were used to train the LV segmentation and the remaining 20 patients with 400 discrete slices were used for testing our method. An interactive graph cut algorithm was utilized reliably to annotate the LV reference standard that was further confirmed by cardiologists. Online data augmentation was performed in the training process to improve the generalization and robustness of our method.

RESULTS : Compared with the segmentation results from the original U-Net and FC-DenseNet56 with Dice similarity coefficient (DSC) of 0.878±0.230 and 0.897±0.189, respectively, our method demonstrated higher segmentation accuracy and robustness for varying LV shape, size, and contrast, achieving DSC of 0.927±0.139. Without online data augmentation, our method resulted in inferior performance with DSC of 0.911±0.170. In addition, compared with the provided results from other existing studies in the LV segmentation of cardiac CT images, our method achieved a competitive performance for the LV segmentation.

CONCLUSIONS : The proposed 8-layer residual U-Net with deep supervision accurately and efficiently segments the LV in CCTA scans. This method has potential advantages to be a reliable segmentation method and useful for the evaluation of cardiac function in the future study.

Li Changling, Song Xiangfen, Zhao Hang, Feng Li, Hu Tao, Zhang Yuchen, Jiang Jun, Wang Jianan, Xiang Jianping, Sun Yong

2020-Nov-26

Cardiac CT angiography, Deep learning, Deep supervision, Left ventricle segmentation, Residual U-Net

Cardiology Cardiology

Cardiac magnetic resonance image diagnosis of hypertrophic obstructive cardiomyopathy based on a double-branch neural network.

In Computer methods and programs in biomedicine

OBJECTIVE : Cardiac magnetic resonance (CMR) imaging is a well-established technique for diagnosis of hypertrophic obstructive cardiomyopathy (HOCM) and evaluation of cardiac function, but the process is complicated and time consuming. Therefore, this paper proposes a cardiomyopathy recognition algorithm using a multi-task learning mechanism and a double-branch deep learning neural network.

METHOD : We implemented a double-branch neural network CMR-based HOCM recognition algorithm. Compared with the traditional classification algorithms such as the ResNet, DenseNet network, contrast the accuracy of network classification of cardiomyopathy is higher by 10.11%.

RESULT : The loss curve of the algorithm basically converges in 100 rounds, and the convergence speed of the algorithm is twice that of the traditional algorithm. The accuracy of this algorithm to classify cardiomyopathy is 96.79%, and the sensitivity is 95.24%, which is 10.11% higher than the conventional algorithm.

CONCLUSION : The CMR imaging automatic recognition algorithm for HOCM capture static morphological and motion characteristics of the heart, and comprehensively enhances recognition accuracy when the sample size is limited.

You Yuanbing, Viktorovich Lysenko Andrey, Qiu Jiawei, Nikolaevich Kosenkov Alexander, Vladimirovich Belov Yuri

2020-Nov-28

Cardiac magnetic resonance imaging, Cardiomyopathy recognition, Deep learning, Double-branch Network, Hypertrophic obstructive cardiomyopathy

Radiology Radiology

Radiomics-based machine-learning method for prediction of distant metastasis from soft-tissue sarcomas.

In Clinical radiology

AIM : To construct and validate a radiomics-based machine-learning method for preoperative prediction of distant metastasis (DM) from soft-tissue sarcoma.

MATERIALS AND METHODS : Seventy-seven soft-tissue sarcomas were divided into a training set (n=54) and a validation set (n=23). The performance of three feature selection methods (ReliefF, least absolute shrinkage and selection operator [LASSO], and regularised discriminative feature selection for unsupervised learning [UDFS]) and four classifiers, random forest (RF), logistic regression (LOG), K nearest neighbour (KNN), and support vector machines (SVMs), were compared for predicting the likelihood of DM. To counter the imbalance in the frequencies of DM, each machine-learning method was trained first without subsampling, then with the synthetic minority oversampling technique (SMOTE). The performance of the radiomics model was assessed using area under the receiver-operating characteristic curve (AUC) and accuracy (ACC) values.

RESULTS : The performance of the LASSO and SVM algorithm combination used with SMOTE was superior to that of the algorithm combination alone. The combination of SMOTE with feature screening by LASSO and SVM classifiers had an AUC of 0.9020 and ACC of 91.30% in the validation dataset.

CONCLUSION : A machine-learning model based on radiomics was favourable for predicting the likelihood of DM from soft-tissue sarcoma. This will help decide treatment strategies.

Tian L, Zhang D, Bao S, Nie P, Hao D, Liu Y, Zhang J, Wang H

2020-Dec-05