AivexaNewsSearch
AI news for builders and product teamsChecked every hour

Google's AI detects high blood pressure from a regular smartwatch without a cuff

Collected Nov 7, 2026

Google Research has developed an AI model that can flag signs of high blood pressure using a standard smartwatch, relying only on the pulse and motion sensors already built into modern watches and requiring no cuff for calibration. The work was described in a Google blog post and the study was published in Nature. According to Google, the capability has already been folded into the blood pressure trends feature in the Google Health app.

The method uses a neural network trained on roughly 500,000 hours of sensor data. Rather than learning from labeled diagnoses, the system was trained by comparing short segments of sensor data and working out which ones came from the same person, allowing it to pick up typical pulse and heartbeat patterns on its own. To produce a result, the system averages those patterns over several days and calculates a probability that the wearer has high blood pressure.

The approach was evaluated in a study of 196 people across five US research sites. In that group the watch performed about as well as a traditional cuff: it identified 65.8 percent of participants with high blood pressure, compared with 55.3 percent for the standard in-office measurement.

The research fits alongside Google's broader health modeling work; the company notes that its SensorFM health model also learns from smartwatch data. In practice, the technique is interesting because it sidesteps the usual calibration requirement for cuffless blood pressure sensing. Many wearable approaches depend on an occasional cuff reading to anchor their estimates, which limits who can use them. Learning general pulse and motion signatures instead suggests a path to screening that works straight out of the box, though the reported detection rates indicate it would function as a prompt to seek a check rather than a measurement tool.

Google is explicit that the method remains a research project and does not replace a formal diagnosis. The stated aim is to nudge people toward getting checked, a point underscored by the company's figure that more than half of people with high blood pressure are unaware of it. This suggests the near-term product role is awareness and triage: surfacing a trend and encouraging a clinical visit, with confirmation still happening through conventional measurement. A likely trade-off is that averaging patterns across several days introduces lag and depends on consistent wear, and the modest sample and detection gap over in-office checks mean the feature is better read as a signal than a substitute for a diagnosis.

Why it matters: Developers and product teams building on wearable health data now have a concrete example of cuffless hypertension screening reaching a consumer app, which may shape expectations for what smartwatch sensors can support. The reported 65.8 percent detection rate, ahead of the 55.3 percent in-office figure in this study, is promising but not diagnostic-grade, so products are likely to position such features as prompts for clinical follow-up. Teams working in health apps should expect regulatory and messaging scrutiny around claims, since the underlying method is still framed as research.

Read at The Decoder

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

Google Research built an AI that spots signs of high blood pressure using only the pulse and motion sensors in a standard smartwatch, no cuff needed. In a 196-person study, its 65.8 percent detection rate beat the traditional in-office measurement. The method is already live in the Google Health app, though it doesn't replace a formal diagnosis. The article Google's AI detects high blood pressure from a regular smartwatch without a cuff appeared first on The Decoder .