SIGMA-PPG: Statistical-prior Informed Generative Masking Architecture for PPG Foundation Model

2026-01-28

AI Summary

This work proposes SIGMA-PPG, a generative foundation model for photoplethysmography (PPG) signals that addresses the challenges of modeling redundancy and noise in PPG data, as well as the limitations of existing masked modeling approaches—namely their tendency to converge to trivial solutions and the insufficient morphological fidelity in contrastive methods. SIGMA-PPG innovatively integrates statistical priors with a reinforcement learning–driven adversarial masking mechanism and incorporates vector quantization to enforce semantic consistency. This design significantly enhances the robustness and semantic density of physiological waveform representations. Pretrained on over 120,000 hours of large-scale PPG data, SIGMA-PPG demonstrates substantially superior average performance across twelve downstream tasks compared to five state-of-the-art baseline models.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Current foundation model for photoplethysmography (PPG) signals is challenged by the intrinsic redundancy and noise of the signal. Standard masked modeling often yields trivial solutions while contrastive methods lack morphological precision. To address these limitations, we propose a Statistical-prior Informed Generative Masking Architecture (SIGMA-PPG), a generative foundation model featuring a Prior-Guided Adversarial Masking mechanism, where a reinforcement learning-driven teacher leverages statistical priors to create challenging learning paths that prevent overfitting to noise. We also incorporate a semantic consistency constraint via vector quantization to ensure that physiologically identical waveforms (even those altered by recording artifacts or minor perturbations) map to shared indices. This enhances codebook semantic density and eliminates redundant feature structures. Pre-trained on over 120,000 hours of data, SIGMA-PPG achieves superior average performance compared to five state-of-the-art baselines across 12 diverse downstream tasks. The code is available at https://github.com/ZonghengGuo/SigmaPPG.

Problem

• photoplethysmography
• foundation model
• signal redundancy
• noise
• masked modeling

Innovation

• Generative Foundation Model
• Prior-Guided Adversarial Masking
• Statistical Priors
• Vector Quantization
• Photoplethysmography (PPG)