TY  - JOUR
AU  - Ho, Thao Thi
AU  - Park, Jongmin
AU  - Kim, Taewoo
AU  - Park, Byunggeon
AU  - Lee, Jaehee
AU  - Kim, Jin Young
AU  - Kim, Ki Beom
AU  - Choi, Sooyoung
AU  - Kim, Young Hwan
AU  - Lim, Jae-Kwang
AU  - Choi, Sanghun
PY  - 2021
DA  - 2021/1/28
TI  - Deep Learning Models for Predicting Severe Progression in COVID-19-Infected Patients: Retrospective Study
JO  - JMIR Med Inform
SP  - e24973
VL  - 9
IS  - 1
KW  - COVID-19
KW  - deep learning
KW  - artificial neural network
KW  - convolutional neural network
KW  - lung CT
AB  - Background: Many COVID-19 patients rapidly progress to respiratory failure with a broad range of severities. Identification of high-risk cases is critical for early intervention. Objective: The aim of this study is to develop deep learning models that can rapidly identify high-risk COVID-19 patients based on computed tomography (CT) images and clinical data. Methods: We analyzed 297 COVID-19 patients from five hospitals in Daegu, South Korea. A mixed artificial convolutional neural network (ACNN) model, combining an artificial neural network for clinical data and a convolutional neural network for 3D CT imaging data, was developed to classify these cases as either high risk of severe progression (ie, event) or low risk (ie, event-free). Results: Using the mixed ACNN model, we were able to obtain high classification performance using novel coronavirus pneumonia lesion images (ie, 93.9% accuracy, 80.8% sensitivity, 96.9% specificity, and 0.916 area under the curve [AUC] score) and lung segmentation images (ie, 94.3% accuracy, 74.7% sensitivity, 95.9% specificity, and 0.928 AUC score) for event versus event-free groups. Conclusions: Our study successfully differentiated high-risk cases among COVID-19 patients using imaging and clinical features. The developed model can be used as a predictive tool for interventions in aggressive therapies. 
SN  - 2291-9694
UR  - http://medinform.jmir.org/2021/1/e24973/
UR  - https://doi.org/10.2196/24973
UR  - http://www.ncbi.nlm.nih.gov/pubmed/33455900
DO  - 10.2196/24973
ID  - info:doi/10.2196/24973
ER  -