Can artificial intelligence predict COVID-19 mortality?

ABSTRACT

OBJECTIVE: COVID-19 infection rapidly spread across the globe and evolved into a pandemic. Artificial intelligence (AI) has been used to predict the spread of the virus and diagnose and treat the disease. In this study, we aimed to predict whether patients admitted to the intensive care unit (ICU) due to COVID-19 infection will result in mortality. PATIENTS AND METHODS: Ninety parameters were used for each 589 ICU patient. The nine parameters with the highest effect on mortality were determined. Four hundred seventy-one patients were used to train the AI with these nine parameters. AI was tested with 118 patient data. RESULTS: AI estimated mortality with 83% sensitivity, 84% specificity, 84% accuracy, and 0.81 F1 score. Therefore, the area under the curve was calculated as 0.91. The results indicate that mortality among COVID-19 patients admitted to the ICU can be predicted based on their laboratory parameters on the first day. CONCLUSIONS: These findings underscore the potential benefits of utilizing AI in managing pandemics.