Detecting signs of hypertension using wearable sensors
This research demonstrates how self-supervised machine learning can turn standard consumer wearables into calibration-free medical screening tools, informing features in the Google Health app.
What happened
Google Research published a study in Nature detailing AI-PPG-ACC-HTN, a multimodal AI system that uses consumer smartwatch optical and motion sensors to passively detect hypertension without requiring cuff calibration. The system uses self-supervised contrastive learning to aggregate physiological signals over multiple days, achieving accuracy comparable to traditional cuff-based clinical workflows when tested against 24-hour ambulatory blood pressure monitoring.
Who it’s about
- Google Research
- AI-PPG-ACC-HTN
- Nature
- Harvard Medical School
- Beth Israel Deaconess Medical Center
Source
Google Research
Detecting signs of hypertension using wearable sensors
More on health
Get a brief picked for you
Sign in to follow topics and companies, save stories and search the archive.
Sign in