The mRhythm Study, an initiative of Cardiogram, the startup involved in providing a workable roadmap to the heart rate data of Apple Watch, has shown that atrial fibrillations can be detected with 97 percent accuracy through algorithms.

The study results have highlighted the unique capabilities of common wearable trackers like smartwatches, to assess, monitor and create newer medical therapeutic possibilities for atrial fibrillation without active patient involvement. Though screening through mobile technologies cannot be a substitute for conventional assessment methodologies, the innovation can raise the success rate of screening those at an enhanced risk and reduce the number of AF cases which remain undiagnosed.

Affecting more than 2.7 million Americans, Atrial Fibrillation (AF) can be defined as common heart arrhythmia. Health startups like Cardiogram and smartphone ECG Company AliveCor have been doing pioneering work in simplifying AF detection. It has often been noticed that AF can trigger strokes though they are often asymptomatic.     

Coordinating with the Health eHeart Study of UCSF, Cardiogram has created the algorithm and enrolled 6,158 users through the Cardiogram Apple Watch app. The collection of crucial data like mobile ECGs and heart rate from enrolled patients was utilized to train a deep neural network. An Apple watch and a 12-lead ECG was used by a group of 51 patients before and after a procedure called cardioversion, used for restoring heart rhythms after an arrhythmia. AF was assessed with 97 percent accuracy, 98 percent sensitivity and 90.2 percent specificity by applying the neural network-derived algorithm.