Hybrid Modeling of Photoplethysmography for Non-invasive Monitoring of Cardiovascular Parameters

2025-11-18

AI Summary

This study addresses the clinical challenge of noninvasive photoplethysmography (PPG)-based monitoring of cardiac output (CO) and stroke volume (SV), overcoming limitations of existing methods that rely on invasive arterial pressure waveforms (APW) and scarce labeled data. We propose a hybrid modeling framework integrating hemodynamic simulation with unlabeled clinical PPG signals, innovatively combining conditional variational autoencoders (CVAEs) and conditional density estimation to achieve interpretable, uncertainty-aware mapping from PPG to time-varying cardiovascular parameters. The method requires neither invasive measurements nor large-scale annotated datasets. Experimental results demonstrate superior accuracy in tracking dynamic CO/SV variations and improved detection of physiological fluctuations compared to supervised learning baselines. This work establishes a novel paradigm for noninvasive, continuous, and personalized cardiac function assessment.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Multi-modal Vision
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Continuous cardiovascular monitoring can play a key role in precision health. However, some fundamental cardiac biomarkers of interest, including stroke volume and cardiac output, require invasive measurements, e.g., arterial pressure waveforms (APW). As a non-invasive alternative, photoplethysmography (PPG) measurements are routinely collected in hospital settings. Unfortunately, the prediction of key cardiac biomarkers from PPG instead of APW remains an open challenge, further complicated by the scarcity of annotated PPG measurements. As a solution, we propose a hybrid approach that uses hemodynamic simulations and unlabeled clinical data to estimate cardiovascular biomarkers directly from PPG signals. Our hybrid model combines a conditional variational autoencoder trained on paired PPG-APW data with a conditional density estimator of cardiac biomarkers trained on labeled simulated APW segments. As a key result, our experiments demonstrate that the proposed approach can detect fluctuations of cardiac output and stroke volume and outperform a supervised baseline in monitoring temporal changes in these biomarkers.

Problem

• Estimating cardiovascular biomarkers from non-invasive PPG signals instead of invasive arterial pressure waveforms
• Addressing scarcity of annotated PPG measurements for predicting cardiac output and stroke volume
• Developing hybrid model combining hemodynamic simulations with unlabeled clinical data

Innovation

• Hybrid model combining hemodynamic simulations with unlabeled data
• Conditional variational autoencoder trained on paired PPG-APW data
• Conditional density estimator for cardiac biomarkers from APW