- A new study from University of California, San Francisco, researchers shows that the Apple (NASDAQ:AAPL) Watch can detect atrial fibrillation with high accuracy – but only when wearers were at rest.
- Atrial fibrillation is a type of irregular heartbeat that can cause blood clots and strokes when untreated and diagnosis is difficult as symptoms often don’t become obvious until too late. An estimated 2.7M Americans and 34M people globally have the problem.
- Researchers used an app from Cardiogram Inc. to access heart-rate sensor data in the Watch. A deep neural network analyzed the heart rate data and step-count data, which was used to determine whether a person was active (explaining a sudden heart rate spike) or at rest.
- The resulting algorithm was able to tell the difference between atrial fibrillation and normal heartbeats with about 97% compared to electrocardiograms. This data came from about 50 patients with atrial fibrillation who were undergoing procedures to restore their heart rates to normal.
- In a test of 1,620 people using the Watch in their everyday lives, the algorithm’s accuracy dropped to about 72% accuracy compared to participant reports of whether or not they had atrial fibrillation.
- Previously: New iPhone X could cost Apple 10% less; Iovine stepping back from Music (March 21)
- Now read: Apple Must Change To A Hands-On Display (And Other Critical Component Strategies)
Original article