Personalized Continuous Blood Pressure Tracking through Single channel PPG in Wearable Scenarios

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Physiological Monitoring

Background:

• Real-time tracking of human physiological states enhances personalized healthcare.
• Photoplethysmography (PPG) is a non-invasive method for monitoring blood flow, blood pressure (BP), and cardiovascular status.
• Continuous BP monitoring via PPG faces challenges like poor signal quality, computational complexity, and calibration needs in wearable applications.

Purpose of the Study:

• To develop a personalized continuous BP tracking pipeline for wearable devices.
• To address challenges in PPG-based BP monitoring, including signal quality, computational load, and calibration.
• To enable robust and long-term BP monitoring across diverse real-world scenarios.

Main Methods:

• Implemented an automatic PPG signal quality grading module to simplify model fitting.
• Introduced a lightweight BP prediction model (SCI-GTCN) to reduce computational complexity.
• Employed an adaptive calibration strategy for sustained BP monitoring performance.

Main Results:

• The pipeline was validated on 134 subjects across various monitoring conditions (daytime, nighttime, abnormal states).
• Performance was assessed during rapid BP changes, circadian rhythm fluctuations, and long-term monitoring.
• Achieved Mean Error ± Standard Deviation (ME±SD) of 0.99±7.91/0.36±5.43 mmHg, meeting AAMI standards.

Conclusions:

• The proposed method demonstrates good robustness and applicability for continuous BP monitoring.
• The pipeline is suitable for deployment on wearable devices, offering a promising solution for the healthcare field.
• The system effectively addresses key limitations of current PPG-based BP monitoring technologies.