THE Apple Watch has been found to detect a heart condition that affects some 2.7 million people in the US, a new study has revealed.
By pairing the smartwatch’s heart rate sensors with artificial intelligence, researchers developed an algorithm capable of distinguishing an irregular heartbeat, known as atrial fibrillation, from a normal heart rhythm – and with 97 percent accuracy.
Atrial fibrillation, although easily treatable, has been difficult to diagnose and the team believes their work could pave the way for new methods to identify the abnormality.
The research was conducted by the University of California, San Francisco, in collaboration with the app Cardiogram, which tracks an individual’s sleep, stress and fitness via the Apple Watch’s built-in hear rate sensor.
The team trained a deep neural network with heart readings from 6,158 Cardiogram users, AppleInsider reported, and recorded 139 million heart rate measurements and 6,338 mobile ECGs from sample group.
The data was then used to develop a unique algorithm to detect the irregular heartbeat that is caused by atrial fibrillation (AF).
AF is a heart condition that can cause blood to collect in the heart, which could result in a clot that leads to a stroke – one in four strokes brought on by AF.
And although the heart abnormality is treatable, it is not easily diagnosed – but the recent study could change that.
To test the algorithm, the team validated it against 51 test patients who were set to receive treatment for AF.
