The Apple Watch, when paired with artificial intelligence and the Cardiogram app, has a high 97-percent accuracy rate when detecting a particular type of abnormal heart rhythm called paroxysmal atrial fibrillation. This discovery was made by the University of California in San Francisco, which based its conclusion in part on data gathered from more than six-thousand volunteers.
The Apple Watch features a heart rate sensor which, though intended for fitness purposes, has proven useful in medical situations before. This time around, researchers conducted the UCSF Health eHeart study using volunteers found via the Cardiogram app. Of these volunteers, 200 reported having the abnormal heart rhythm detailed above, something that had been determined via a proper diagnoses.
Using both their data and the heart rate data of the healthy volunteers, the researchers created a deep learning neural network AI that could detect the presence of paroxysmal atrial fibrillation using the Apple Watch’s heart rate sensor data. The discovery of this condition could save affected individuals from a potential stoke, which is said to result in about a quarter of cases from the atypical heartbeat.
Though you won’t find the Apple Watch in use as an accepted way to detect abnormal heart rhythms, this study does highlight the growing potential of consumer technology products. We’ve seen a number of health-related services and technologies make their way onto wearables and smartphones, including built-in heart rate sensors on certain handsets and things like on-demand access to doctors through services like Samsung Health.