CLEP-GAN: An Innovative Approach to Subject-Independent ECG Reconstruction from PPG Signals

2025-02-24

AI Summary

This work addresses the critical challenge of subject-independent, non-invasive ECG reconstruction from photoplethysmography (PPG) signals—hampered by limited dataset diversity and severe noise-induced poor model generalizability. We propose an ordinary differential equation (ODE)-based synthetic data generation method that produces physiologically interpretable, high-fidelity ECG-PPG paired samples—the first of its kind for principled data augmentation. Further, we design a multi-objective learning framework tailored for subject-independent reconstruction, integrating contrastive learning, generative adversarial network (GAN)-based adversarial training, and attention-gated mechanisms. We systematically characterize the impact of age and sex on reconstruction performance. Experiments demonstrate state-of-the-art performance in cross-subject ECG reconstruction, with significantly improved noise robustness and generalization capability—establishing a new paradigm for high-accuracy, minimally invasive ECG estimation in wearable devices.

Technology Category

• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Computer Vision: Computational Photography, Image & Video Synthesis
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

This study addresses the challenge of reconstructing unseen ECG signals from PPG signals, a critical task for non-invasive cardiac monitoring. While numerous public ECG-PPG datasets are available, they lack the diversity seen in image datasets, and data collection processes often introduce noise, complicating ECG reconstruction from PPG even with advanced machine learning models. To tackle these challenges, we first introduce a novel synthetic ECG-PPG data generation technique using an ODE model to enhance training diversity. Next, we develop a novel subject-independent PPG-to-ECG reconstruction model that integrates contrastive learning, adversarial learning, and attention gating, achieving results comparable to or even surpassing existing approaches for unseen ECG reconstruction. Finally, we examine factors such as sex and age that impact reconstruction accuracy, emphasizing the importance of considering demographic diversity during model training and dataset augmentation.

Problem

• Reconstructing ECG signals from PPG signals
• Enhancing training diversity with synthetic data
• Improving subject-independent ECG reconstruction accuracy

Innovation

• Synthetic ECG-PPG data generation
• Subject-independent PPG-to-ECG model
• Contrastive and adversarial learning integration