The Apple Watch can detect diabetes in previously diagnosed patients with up to 85-percent accuracy. That number comes from Cardiogram in conjunction with the UCSF Department of Medicine, which held a study to understand how useful Apple’s wearable hardware is at detecting diseases.
Cardiogram used Apple Watch hardware with its sophisticated algorithm to come up with results among the test subjects.
Part of a DeepHeart study, Cardiogram studied the data of 14,011 Apple Watch users to see if the Apple Watch’s array of sensors could be used to detect diabetes. Cardiogram used multiple tests, including unsupervised, semi-supervised and heuristic to cross reference heart rate data with the algorithm.
The study was done without the help of any additional medical hardware and yet the Apple Watch was able to successfully diagnose 85-percent of users who had previously been diagnosed with diabetes.
This latest news has the potential to be a huge breakthrough for the disease. About 100 million adults suffer from pre-diabetes and diabetes and more than one in four go undiagnosed, reports the Center for Disease Control and Prevention. The Apple Watch can ago a long way to improve that number.
Cardiogram does say that it cannot make definitive diagnosis using consumer grade heart sensors, like the one found in the Apple Watch. In fact, the end goal of the DeepHeart study was never to successfully diagnose diabetes through an Apple Watch, though the results showed that it could be possible.
All that being said, combining the data obtained from the Apple Watch with Cardiogram’s algorithm has the opportunity to be a game-changer in early detection of diabetes.