​
Invention Summary:
Recently, the data-driven capabilities of deep learning have been leveraged in blood pressure (BP) monitoring systems that utilize photoplethysmogram (PPG) signals. These advancements aim to enhance estimation accuracy, clinical applicability, and ultimately, patient outcomes. However, challenges persist—particularly regarding the generalizability and reliability of these models.
Rutgers researchers introduce a novel system and method for continuous monitoring of significant changes (elevations or drops) in blood pressure using a self-contrastive masking (SCM) multi-stage machine learning model applied to PPG data. The system receives PPG signals from a wearable sensor, processes the data through segmentation, feature extraction, and decision-making phases, and outputs real-time determinations of BP changes. Furthermore, the technology will signal any deviations of BP outside of a provided threshold and generate an alarm to alert the patient. Validation using data from PulseDB and new patient groups shows the system works reliably across different people and situations, needing only PPG input and making it easy to integrate into smart wearables—helping patients get timely alerts during episodes of hypertensive or hypotensive emergencies and supporting more accessible, everyday blood pressure monitoring.
Market Applications:
- Wearable Heath Monitors
- Remote Patient Monitoring
- Research in cardiovascular health and diagnostics
- Hypertensive / Hypotensive Emergency Indicators
Advantages:
- Improved Generalizability: transitioning from directly estimating BP, to detecting over-threshold BP changes, improves the generalizability of the model, as demonstrated by testing on unseen patients.
- High Sensitivity to BP Changes: Capable of detecting acute changes in blood pressure over short intervals, increasing clinical applicability.
- Non-Invasive and Seamless: Utilizes PPG data from widely available and comfortable sensors, enabling continuous and unobtrusive monitoring.
Publications:
- Wang, Weinan et al. “BP-Net: Monitoring "Changes" in Blood Pressure Using PPG With Self-Contrastive Masking.” IEEE journal of biomedical and health informatics vol. 28,12 (2024): 7103-7115. doi:10.1109/JBHI.2024.3422023
Intellectual Property & Development Status: Provisional application filed. Patent pending. Available for licensing and/or research collaboration. For any business development and other collaborative partnerships, contact: marketingbd@research.rutgers.edu
