How a bad day at work led to better COVID predictions

Dr. Sejal Morjaria, like other physicians treating COVID-19, found it hard to predict how her infected cancer patients at Memorial Sloan Kettering Cancer Center (MSK) would react to the virus. She and her husband, CSHL Associate Professor Saket Navlakha, a computer scientist, worked together to develop a machine-learning solution that uses 50 variables available when a patient is first diagnosed to predict how severely their COVID-19 symptoms are likely to progress.