Using machine learning and PHE data to help hospitals cope with COVID-19

ellis 01 April 2020 - 01 April 2020

01 April 2020 • 15:40 - 15:55

Mihaela van der Schaar (University of Cambridge)

Mihaela van der Schaar’s talk highlights how machine learning methods trained on Public Health England CHESS data can be used to predict how COVID-19 will affect resource needs in the hospitals, such as ventilators and ICU beds, at the individual patient level and at the hospital level. She presents the Adjutorium system which is capable of forecasting personalized risk profiles and benefits from resources for each patient, as well as recommend resource allocation schedules based on expected benefit. Adjutorium aims to reliably assist clinicians and hospital managers in making decisions about COVID-19 patients.