Newly published research suggests that the Apple Watch’s built-in suite of sensors can detect abnormal heart rhythms with 97 percent accuracy.
The results of the study, which was a joint venture between researchers at the University of California, San Francisco and the team behind heart monitoring app Cardiogram, were published in JAMA Cardiology on Wednesday. The research confirms the results of a preliminary study published back in May 2017.
The researchers collected more than 139 million heart rate and step count measurements between February 2016 and March 2017. The data was pulled from 9,750 participants who enrolled in UCSF’s Health eHeart Study and had downloaded the Cardiogram app, and from 51 patients undergoing cardioverison at UCSF, the study states.
All of the data was then used to train DeepHeart, Cardiogram’s proprietary deep neural network, to detect abnormal heart rhythms.
What they found is that DeepHeart was able to differentiate between normal heart rhythms and atrial fibrillation (AFib) with a 97 percent accuracy rate — both when they tested heart patients and Cardiogram users with no known heart issues.
Interestingly, the study suggests that the Apple Watch alone — with its built-in suite of sensors — can detect AFib more accurately than an accessory recently approved by the FDA. That accessory, a third-party Apple Watch strap called KardiaBand, was able to detect AFib with about 90 percent accuracy.
Atrial fibrillation, also known as an abnormal heart rhythm, can increase the risk of everything from heart attack and stroke to dementia. At times, people can exhibit little or no symptoms — which can make detection of the condition difficult.
That’s where the Apple Watch and similar wearables can help. While they likely won’t replace EKGs or other traditional medical systems anytime soon, they can alert wearers to potential problems. In other words, heart rate wearables won’t be able to diagnose AFib — but they could help users get the treatment they need much earlier.
Cardiogram and UCSF researchers have previously carried out studies using DeepHeart to determine if the Apple Watch can detect other conditions and diseases. They found that the Apple wearable can detect diabetes with 85 percent accuracy, hypertension with 82 percent accuracy, and sleep apnea with 92 percent accuracy.
And while Cardiogram and UCSF exclusively use the Apple Watch, the results bode well for basically any wearable with similar heart rate monitors and sensors.
Apple has been increasingly interested in the health sector in recent years, and the Cupertino tech giant recently launched its own study to see if its flagship wearable can detect AFib. The so-called Apple Heart Study, conducted in partnership with researchers at Stanford, officially kicked off last November.
Participants in that study will be contacted by researchers if their wearables detect an abnormal heart rhythm. From there, researchers will ask the user to wear an ePath monitor to further test heart health.
The Apple Heart Study is open to any Apple Watch owner who wishes to participate. Interested users can sign up by downloading and installing the Apple Heart Study app and following the instructions.