Aortic Valve Disease Detection from PPG via Physiology-Informed Self-Supervised Learning

2026-02-04

AI Summary

This study addresses the challenges of aortic valve disease screening, which are hindered by the high cost of echocardiography and the scarcity of labeled photoplethysmography (PPG) data. To overcome these limitations, the authors propose a physiology-guided self-supervised learning framework (PG-SSL) that integrates clinical physiological knowledge into PPG morphological phenotyping for the first time. The method formulates pulse pattern recognition as a proxy task, pre-training on 170,000 unlabeled PPG samples and fine-tuning with minimal labeled data using a dual-branch gated fusion architecture. The approach achieves AUCs of 0.765 and 0.776 for detecting aortic stenosis and regurgitation, respectively, significantly outperforming supervised baselines. Moreover, the model’s outputs are validated as digital biomarkers with independent prognostic value, offering an effective solution to the label scarcity problem in medical AI.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Search and Optimization: Learning to Search
• Computer Vision: Multi-modal Vision

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling

Abstract

Traditional diagnosis of aortic valve disease relies on echocardiography, but its cost and required expertise limit its use in large-scale early screening. Photoplethysmography (PPG) has emerged as a promising screening modality due to its widespread availability in wearable devices and its ability to reflect underlying hemodynamic dynamics. However, the extreme scarcity of gold-standard labeled PPG data severely constrains the effectiveness of data-driven approaches. To address this challenge, we propose and validate a new paradigm, Physiology-Guided Self-Supervised Learning (PG-SSL), aimed at unlocking the value of large-scale unlabeled PPG data for efficient screening of Aortic Stenosis (AS) and Aortic Regurgitation (AR). Using over 170,000 unlabeled PPG samples from the UK Biobank, we formalize clinical knowledge into a set of PPG morphological phenotypes and construct a pulse pattern recognition proxy task for self-supervised pre-training. A dual-branch, gated-fusion architecture is then employed for efficient fine-tuning on a small labeled subset. The proposed PG-SSL framework achieves AUCs of 0.765 and 0.776 for AS and AR screening, respectively, significantly outperforming supervised baselines trained on limited labeled data. Multivariable analysis further validates the model output as an independent digital biomarker with sustained prognostic value after adjustment for standard clinical risk factors. This study demonstrates that PG-SSL provides an effective, domain knowledge-driven solution to label scarcity in medical artificial intelligence and shows strong potential for enabling low-cost, large-scale early screening of aortic valve disease.

Problem

• Aortic Valve Disease
• Photoplethysmography
• Label Scarcity
• Early Screening
• Self-Supervised Learning

Innovation

• Physiology-Guided Self-Supervised Learning
• PPG morphological phenotypes
• proxy task
• gated-fusion architecture
• digital biomarker