Deriving Endotypes of Critically Ill COVID-19 Patients and Predicting Impending Cardiovascular Collapse

Soojin Park, is an Associate Professor of Neurology at Columbia University College of Physicians and Surgeons

We propose to learn endotypes of the critically ill COVID-19 patient population using a combination of overt and latent features from observable characteristics, mechanistic information, and response to therapy. We will also featurize continuous cardiopulmonary telemetry data and develop a deep learning model for use as a sequential classifier for impending cardiovascular collapse.