What Google Research did
Google Research has developed an artificial intelligence system capable of recognizing signs of high blood pressure using only the wrist and motion sensors in an ordinary smartwatch, without the need for a cuff-based monitor for calibration. The Decoder reports this, citing a Google blog post and the study published in Nature.
The neural network was trained by comparing short segments of data collected by the sensors, totaling almost 500,000 hours, and learning to recognize which belonged to the same person. In this way, the system identified typical patterns of the wrist and heartbeat without being guided by diagnoses. The readings thus obtained are averaged over several days until a probability of hypertension is calculated.
The study's results
The trial involved 196 people at five research centers in the United States. According to the reported data, the watch identified 65.8% of people with high blood pressure, versus 55.3% for standard measurement performed in a clinic. The comparison therefore concerns not the absolute accuracy of the device but its ability to detect cases compared with traditional clinical practice.
Google says the results have already been incorporated into the blood pressure trends feature in the Google Health app. The method, however, remains a research project and does not replace a formal diagnosis: the stated goal is to encourage people to undergo checkups. According to the company, more than half of people with high blood pressure do not know they have it.
What is confirmed and what remains to be clarified
The material confirms three facts: the existence of the study in Nature, the sample size and the reported detection rates. The claim that the feature is already active in the Health app, however, is a Google statement, not a fact independently verifiable from the material.
It is also not specified how the calculated probability is presented to the user, how often the data are updated, or what limitations the system has in populations other than the one studied. The sample of 196 people remains small relative to the potential reach of a tool of this kind, and the material does not indicate whether further validations are planned. The defensible conclusion is that the technology aims at a healthcare triage use, not a diagnostic one.
