Continuous Non-Invasive Blood Pressure Measurement Using 60 GHz-Radar-A Feasibility Study.

Area of Science:

• Biomedical Engineering
• Cardiovascular Health Technology
• Signal Processing

Background:

• Current upper-arm cuff sphygmomanometers offer static blood pressure readings, lack continuous monitoring capabilities, and can be inaccurate and uncomfortable.
• Accurate and continuous blood pressure monitoring is crucial for assessing cardiovascular health and managing hypertension.

Purpose of the Study:

• To develop and evaluate a radar-based system for extracting blood pressure information from arterial pulse wave signals.
• To assess the feasibility of using a neural network model with radar-derived features for blood pressure estimation.

Main Methods:

• A radar system was employed to detect skin movements caused by artery pulsations, extracting pressure waves.
• 21 distinct features were extracted from the pressure waves and combined with demographic data (age, gender, height, weight).
• A neural network regression model was trained using data from 55 subjects, with 126 networks analyzed for predictive performance.

Main Results:

• A shallow neural network (two hidden layers) achieved a systolic blood pressure error of 9.2±8.3 mmHg and a diastolic error of 7.7±5.7 mmHg.
• The model's performance did not meet AAMI and BHS standards, but optimization was not the primary study objective.
• The radar-based approach demonstrated significant potential in capturing blood pressure variations using the extracted features.

Conclusions:

• The developed radar-based method shows promise for non-invasive, continuous blood pressure monitoring, potentially for integration into wearable devices.
• Further improvements to the approach are needed to meet clinical accuracy standards for home use or screening applications.
• This technology offers a potential alternative to traditional cuff-based methods for long-term cardiovascular health assessment.