World first for AI and machine learning to treat Covid patients worldwide

In what’s known as federated learning, the research applied an algorithm to analyse anonymised electronic patient health data and chest x-rays from 10,000 Covid patients worldwide, including 250 at Addenbrooke’s Hospital.

The study – dubbed EXAM – took just two weeks of AI ‘learning’ to achieve high-quality predictions on how much extra oxygen a patient would need in the first days of hospital care.

To maintain strict patient confidentiality, the patient data was fully anonymised and an algorithm was sent to each hospital so no data was shared or left its location.

Once the algorithm had ‘learned’ from the data, the analysis was brought together to build an AI tool which could predict the oxygen needs of hospital Covid patients anywhere in the world.

This model can be used to help frontline physicians worldwide.

This is an abridged version of the article first published on our website on 15 September 2021.