But the scientist revealed that detection can only be done when the user is sleeping on his back. The study, conducted using step-count and heart rate sensors at Apple Watch to monitor data related to a heart condition called atrial fibrillation.
Atrial fibrillation is an unusual heartbeat condition, which can trigger blood clots and strokes. The data studied by the scientists was collected from 6,680 Apple Watch owners, including 50 owners who suffered from atrial fibrillation.
The data is accessed via an application from Cardiogram Inc, a company that also backs the fund for this research. The information was then analyzed more deeply with the help of the machine learning network, and produced some interesting conclusions.
First, Apple Watch diagnoses atrial fibrillation with an accuracy of 97 percent over a prominent cardiogram, as long as the patient is resting and not doing physical activity. In the case of a patient moving or performing a regular daily task, the detection rate decreased to 72 percent.
This is the difference between the heart rate pulse rate, causing Apple Watch less able to detect heart rate for a day. Atrial fibrillation affects more than 2.7 million Americans and a total of 34 million people worldwide.
The research is part of the eHealth Heart research project that began in 2013, and aims to prevent heart disease by utilizing the mobile technology available today. To date, the project has 160 thousand participants, and is targeted to reach one million participants by the end of 2018.