Google Research detects hypertension via smartwatch sensors

It combines PPG and accelerometer signals, learns physiological features via self-supervised pre-training, and aggregates daily data with height to predict hypertension probability. In a multi-center clinical study, it achieved 65.8% sensitivity and 90.0% specificity against ABPM gold standard, outperforming office-based screening in positive predictive value.

Summarised from the original article by an open-source model, then checked automatically: every number and English name must appear in the source. The original article is authoritative.