Analyzing data from wearable devices using artificial intelligence (AI) may predict a high blood pressure crisis up to three days before it occurs, potentially giving patients and physicians more time to intervene.
These findings were presented today at the American Heart Association's Scientific Sessions 2026.
An AI tool then analyzed 47 features, including blood pressure fluctuations, changes during sleep, morning blood pressure increases, heart rate and heart rate variability, to predict a hypertensive crisis within the next 72 hours.
Over the course of the study, 312 participants experienced 418 hypertensive crises.
The AI model was trained on 80% of the participants and validated on the remaining 20%, achieving high accuracy when discriminating between patients at higher and lower risk.
【MORE】