Google researchers have developed a system that passively measures heart rate and resting heart rate through a smartphone’s front-facing camera.
The system, called passive heart rate monitoring, or PHRM, captures eight-second facial video clips in the moments after a user unlocks their phone via face recognition. An on-device deep learning pipeline then estimates heart rate from subtle changes in how light interacts with the skin as blood pulses through it, aggregating measurements throughout the day into a daily resting heart rate estimate.
Google built PHRM using more than 350,000 video clips from nearly 700 consented research participants in laboratory and free-living settings, per a June 1 study in Nature. The system measured heart rate within 10% of ECG-derived readings across light, medium and dark skin tones — the only model among 16 tested to meet that accuracy bar for all groups — and its daily resting heart rate estimates landed within 4.39 beats per minute of a Fitbit Charge 6 on average.
With roughly 5 billion people owning smartphones, the researchers said the approach could “democratize the benefits of heart health tracking” for populations less likely to adopt wearables. Future studies could test the technology for clinical uses such as monitoring patients with atrial fibrillation or heart failure, the researchers said, and Google is releasing the dataset and a pretrained model to qualified researchers.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.