Towards passive heart health monitoring via smartphone camera

Details

Google Research has published a passive heart rate monitoring system (PHRM) in Nature that uses the front-facing camera of a smartphone to measure heart rate and resting heart rate in the background during everyday use. Triggered by face unlock events, the system applies deep learning to estimate HR with a mean absolute percentage error (MAPE) below 10% — meeting industry accuracy standards across all skin tones — and estimates daily resting heart rate with a mean absolute error under 5 bpm, matching dedicated wearables like Fitbit. The research builds on prior Google work using a finger over the rear camera for on-demand HR measurement and studies linking that signal to cardiovascular disease prediction. The key advance here is fully passive measurement: no user action required, no wearable needed. With the publication, Google is releasing the largest and most diverse public dataset of smartphone facial videos for health research, along with a pre-trained PHRM-mini model. Qualified researchers can apply for access. The clinical implications are significant: roughly five billion people own smartphones with front-facing cameras, making this a potential path to cardiovascular monitoring at a scale no wearable ecosystem can match. FDA classification and practical deployment pathways remain unaddressed in this paper.