CHICAGO — According to a study presented at the Heart Rhythm Society Annual Scientific Sessions, the heart rate sensor of the Apple Watch can detect symptomless atrial fibrillation when paired with artificial intelligence.
Using photoplethysmography, smartwatches can detect heart rate and potentially provide a way to detect silent AF, the researchers wrote in an abstract.
The researchers used the smartwatch’s photoplethysmographic sensors in combination with a deep learning algorithm to determine if AF could be detected using heart rate data.
“Our results show that common wearable trackers like smartwatches present a novel opportunity to monitor, capture and prompt medical therapy for AF without any active effort from patients,” Gregory M. Marcus, MD, MAS, endowed professor of atrial fibrillation research and director of clinical research for the division of cardiology at University of California, San Francisco, said in a press release.
Gregory M. Marcus
The study enrolled patients with AF without a ventricular paced rhythm who were undergoing cardioversion.
According to the release, data from 6,158 users of Cardiogram for Apple Watch were used to develop a “deep neural network” algorithm to distinguish AF from sinus rhythm.
“One of the major advantages of this approach is that it leverages devices that individuals are purchasing on their own and wearing continuously anyway,” Jose Sanchez, MD, cardiac electrophysiology fellow at University of California, San Francisco, told Cardiology Today.
Sanchez, Marcus and colleagues validated the algorithm in a cohort of 51 patients with AF scheduled to undergo cardioversion. Each of the 51 patients (mean age, 66 years; 68% men; 61% white; 54% with normal left ventricular function) who underwent successful cardioversion wore an Apple Watch 20 minutes before and 20 minutes after cardioversion; performance of the deep learning algorithm was compared with that of a 12-lead ECG.
Compared to previously validated algorithms for AF, the C-statistic for AF detection using the deep learning algorithm was significantly higher (0.97), as was the sensitivity of 98.04% and the specificity of 90.2%, the researchers reported.
“While mobile technology screening won’t replace more conventional monitoring methods, it has the potential to successfully screen those at an increased risk and lower the number of undiagnosed cases of AF,” Marcus said in the release. – by Dave Quaile
Sanchez JM, et al. AF detection and ablation outcomes: Answering questions that matter to patients. Presented at: Heart Rhythm Society Annual Scientific Sessions; May 10-13, 2017; Chicago.
Disclosure: Marcus reports receiving research grants from Cardiogram Inc. Sanchez reports no relevant financial disclosures.