Wearable tests and AI to detect infectious cases

The Department of Defence is currently testing if wearable health trackers and artificial intelligence (AI) can detect signs of the virus early so as to catch it before it spreads among soldiers. The Rapid Analysis of Threat Exposure (RATE) clinical trial is thus actively recruiting people to test the algorithm. By wearing a Garmin watch and an Oura ring to monitor the vital signs continuously, the AI algorithm analyzes the data and identify changes in heart rates and temperature, which are early signs of the illness.

The project aims to develop a system that can detect any infectious disease in its earlier stage so as to prevent contamination of others. After the spread of the pandemic, researchers decided to focus the algorithm on COVID-19 and to deploy it to track infection among the military.

The AI algorithm can analyze 165 biomarkers to generate a risk score for an individual. It uses simple measurements like temperatures and heart rates. This combination allows to better predict the illness. A risk score is generated every hour for each individual so this continuous testing can establish a baseline for biomarkers and spot changes in these measurements.

To protect the participant’s privacy, the algorithm collects no personal data and the user can sign up through the device to connect to the cloud service. The device then sends the biometric data to the cloud and the individual can check their risk score in total security and privacy. It is thus up to them to report the score.

Researchers use biomarker information from datasets on Physionet.org to train the RATE algorithm and the publicly available data is from various hospitals around the country. All of these combined datasets make it more valuable and efficient.

The algorithm was created by Philips and various other partners. They are looking to expand the project to the Veteran Administration and for civilian use whenever it is possible.

Related Posts