查看文章

The healthcare support system worldwide faced an enormous burden during the COVID-19 pandemic. One major obstacle is the stratification of patients to predict the severity of the disease and identify effective treatment options, as the available patient data is incomplete and limited. We have developed a decision-making tool based on our artificial intelligence (AI) and machine learning (ML) platform, which can impute missing key clinical parameters for predicting potential disease severity outcomes in patients infected with SARS-CoV-2. We have developed an ML-driven patient stratification model integrated into the hybrid platform BIOiSIM™, which can predict COVID-19 disease outcomes from incomplete clinical datasets. Published clinical datasets were used to train and validate classification models against patient survival. Automated statistical algorithms were deployed to select the most essential features from the input descriptors. Missing values of the patient biomarkers were filled by the multivariate imputation technique. ML algorithms were combined with model stacking to improve the overall performance. The results demonstrated that the DL/ML model was successfully implemented to predict COVID-19 disease outcomes with an accuracy of 0.85, AUC ROC score of 0.88, and F1 Score of 0.89 using patient datasets that were missing 75% of parameter values. Our results demonstrate that AI/ML-driven predictions performed by BIOiSIMTM can assist in generating personalized treatment strategies for COVID-19 patients in healthcare units with limited diagnostic capacities.