A new study found the Apple Watch’s heart rate monitor could detect abnormal heartbeats within 97 percent accuracy when paired with software that incorporates artificial intelligence.
University of California San Francisco researchers conducted the study using Cardiogram, a heart rate measurement app, to explore the possibility of the Apple Watch’s detecting a stroke before it happens.
The team started by looking at atrial fibrillation—the most common of irregular heartbeats—since 15 percent are tied to strokes, according to the American Stroke Association.
Out of the 6,158 participants involved in the research, 200 had been diagnosed with an abnormal heartbeat prior to the study. Researchers then developed a deep-learning algorithm and trained it to detect irregularities based on the data gleaned from the Apple Watch.
The findings are preliminary and the study is ongoing since the eventual goal is to detect early warning signs of stroke. But, the results are promising and part of a growing body of research using the device.
Earlier this year, a Johns Hopkins University study using data from the Apple Watch found stress and missed sleep were the most common seizure triggers for epileptics.
This article was featured in the InsideHook newsletter. Sign up now.