Non-Invasive Continuous-Time Blood Pressure Estimation from a Single Channel PPG Signal using Regularized ARX Models

Area of Science:

• Biomedical Engineering
• Cardiovascular Physiology
• Signal Processing

Background:

• Continuous blood pressure (BP) monitoring is crucial for managing hypertension and cardiovascular diseases.
• Wearable optical sensors can capture photoplethysmography (PPG) signals for indirect BP measurement.
• Accurate modeling of the PPG-to-BP relationship is necessary for reliable estimation.

Purpose of the Study:

• To investigate the suitability of autoregressive with exogenous input (ARX) models with kernel-based regularization for continuous BP estimation.
• To evaluate the feasibility of deriving continuous BP from a single PPG signal source.
• To demonstrate a proof-of-concept for non-invasive, continuous BP monitoring.

Main Methods:

• Analysis of 10 PPG time-series from a wearable optical sensor.
• Utilized corresponding blood pressure reference values for model training and validation.
• Applied autoregressive with exogenous input (ARX) models with kernel-based regularization.

Main Results:

• The study demonstrated promising results in continuous BP estimation.
• Feasibility of continuous BP estimation from a single PPG source was shown.
• Successful application of ARX models for PPG-to-BP dynamics during resting states.

Conclusions:

• Autoregressive models with kernel-based regularization show potential for continuous BP monitoring.
• This approach offers a non-invasive method for estimating BP using wearable PPG sensors.
• Further research is warranted to validate these findings in diverse populations and conditions.