Hypertension, the "silent killer" affecting over one billion people globally, remains largely undiagnosed, especially among younger adults aged 18-44. Traditional office BP measurements are the primary screening method, but they only identify hypertension after it's established. This gap is critical because early detection enables life-saving interventions and reduces long-term cardiovascular risks like stroke or heart failure.
Existing smartwatch wearables often require regular cuff calibration, limiting their utility as independent screening tools. In a Nature study, researchers developed AI-PPG-ACC-HTN, a multimodal AI system that passively detects hypertension signs using a consumer smartwatch without any cuff calibration. The system analyzes passive physiological signatures from two sensors: Photoplethysmography (PPG) tracks blood volume changes at the wrist, while Accelerometry (ACC) captures motion-related ballistocardiogram vibrations.
Using a two-stage approach—self-supervised pre-training on 492,347 hours of sensor data from 2,109 participants, followed by person-level aggregation and logistic regression classification—the AI achieves accuracy comparable to traditional clinical workflows for both initial and confirmatory screening. Validated across five US sites with 196 participants, this breakthrough informs the Blood Pressure Trends Health Guardian feature in Google Health.
Source: Google AI Blog · Summarized by HeadlinesBriefing