[Google AI Identifies High Blood Pressure Using Ordinary Smartwatch Signals]
Google researchers and medical collaborators leveraged ordinary smartwatch signals to flag potential high blood pressure without requiring arm‑cuff calibration.
A study published in Nature on November 5 proposes a method for prompting clinical evaluations during routine wear, though the system failed to detect approximately 33% of hypertension cases.
The AI identified 65.8% of participants with hypertension using a clinical reference test (sensitivity), while correctly ruling out hypertension in 90% of those without the condition (specificity).
This prospective, multicenter study followed 196 participants for seven days of smartwatch monitoring and compared outcomes with 24‑hour ambulatory blood‑pressure monitoring.
All participants had no known cardiovascular disease and were not on blood‑pressure medication, which limits the generalization of the findings to patients already receiving treatment.
The 95% confidence intervals were 54%–76.3% for sensitivity and 83.2%–94.7% for specificity.
Two groups, two screening measures
Two groups, two screening measures
With hypertensionWithout hypertension
Of participants with hypertension on 24-hour ambulatory monitoring, 65.8% were detected and 34.2% were missed.
Participants with hypertension65.8% detected; 34.2% missed. These are different denominator groups, not the overall population. PARTICIPANTS WITH HYPERTENSION 65.8%34.2% DetectedMissed Sensitivity: 65.8% | 95% CI: 54.0% to 76.3% Missed share derived as 100% – 65.8%.
Of participants without hypertension on the reference test, 90.0% screened negative and 10.0% produced false alerts.
Participants without hypertension90.0% correctly negative; 10.0% false alerts. These are different denominator groups, not the overall population. PARTICIPANTS WITHOUT HYPERTENSION 90.0%10.0% Correctly negativeFalse alerts Specificity: 90.0% | 95% CI: 83.2% to 94.7% False-alert share derived as 100% – 90.0%.
Screening can flag a reason for a clinical check; a negative result cannot exclude hypertension.
196 participants in a selected, risk-enriched cohort, without known cardiovascular disease or antihypertensive medication. Bars show conditional study rates, not predictive values for all smartwatch users. This is screening research, not a standalone diagnosis.
How everyday sensors become a screening signal
The system, called AI-PPG-ACC-HTN, combines optical pulse-wave measurements with accelerometer signals.
According to Google Research’s explanation, it aggregates those signals across days and combines them with the user’s height to produce a probability of hypertension.
It does not report an exact systolic and diastolic blood-pressure reading. No personal cuff reading is needed to calibrate the smartwatch system, though cuff-based measurements remained the reference the researchers used to test it.
The waveform model was pretrained using 492,347 hours of unlabeled sensor data from 2,109 consenting participants.
AI can’t scale without trust. Trust starts with the data layer
The study supplement describes the prospective cohort as intentionally risk-enriched. In this selected group, 80.6% of positive smartwatch screens matched hypertension on the reference test, the measure known as positive predictive value.
The study supplement describes the prospective cohort as intentionally risk-enriched. In this selected group, 80.6% of positive smartwatch screens matched hypertension on the reference test, the