Study tests if smartwatch biometrics could spot early COVID-19 clues

Originally published in Biometric Update

Fans of siccing biometrics on COVID-19 have another possible solution.

Purdue University is studying the possibility that an AI-enabled smartwatch might be able to spot symptoms thought to be associated with early onset of the infectious disease through biometrics.

The theory is that the wearables could pick up subtle changes in a person’s breathing and heart rate that seem to announce COVID-19 infection. Getting a continuous readout of relevant biometric data could spot patterns worth investigating, according to the Purdue researchers.

But like so much else about this coronavirus, the early promise of this idea could prove ephemeral, which is why the school has formed a study.

An article published by the American Association for the Advancement of Science reports that any encouraging data from the study would be handed to physIQ, Chicago-based digital health company affiliated with Purdue. PhysIQ would plug the trends into new smartwatch algorithms.

That is a big if, and not the only one for biometrics. A Carnegie Mellon University researcher in April pushed the idea that unique audio signatures in one’s cough could help diagnose COVID-19.

Right now, up to 100 people will be studied to see if the smartwatch idea is practical, unobtrusive and easy enough for people to reliably use.

Samsung Galaxy watches loaded with a physIQ application will be mailed to participants. They also will receive adhesive chest biosensors for recording a single-lead electrocardiogram (EKG) signal.

Collected data will be analyzed for usefulness in training algorithms, according to Purdue.