Multi-stage Deep Learning for Early Predicting Severe COVID-19 Patients

Overview

COVID-19 was wide spreading worldwide, which causes huge death toll. Many countries run out of medical resources due to huge amount of emerging cases everyday. However, most of them are mild cases which do not need special treatment. Early intervention on patients who have high risks to develop to severe cases can effectively reduce the death rate. In this project, we aim to predict potential severe cases based on patients' initial diagnosis and early-stage test results, therefore they can be treated as early as possible. We first formulate predicting patient's future condition as a stagewise learning problem and predict patients' future condition n-stage-away based on their testing results at current and previous stages.

Project Framework

Members

Previous members include Dr. Zhuo Li. Thanks for his contributions.