AI-Assisted Home Cardiovascular Diagnostics
Machine learning algorithms for home-based blood pressure assessment from wearable sensors
Hypertension affects nearly half of US adults and is the leading modifiable risk factor for cardiovascular disease. Accurate home blood pressure monitoring is critical for effective management, yet current consumer devices are prone to measurement error from improper technique.
The project. In collaboration with Dr. Julio Lamprea and the UCSF Cardiovascular Care and Prevention Center, we are developing AI-assisted algorithms that assess blood pressure measurement technique and provide real-time feedback — enabling more reliable home monitoring.
Our approach.
- Developed the core algorithmic framework for technique assessment, now protected under a US provisional patent (Application No. 63/984,962, filed February 2026, assigned to The Regents of the University of California)
- Ongoing prospective data collection at the UCSF Cardiovascular Prevention Center for algorithm validation and fine-tuning
- Integration with wearable sensor data for continuous, passive monitoring
Impact. Accurate home blood pressure monitoring has direct clinical implications for hypertension diagnosis, treatment titration, and cardiovascular risk reduction at scale.