New AI model screens aortic valve disease using PPG signals

Researchers have developed a novel Physiology-Guided Self-Supervised Learning (PG-SSL) method to screen for aortic valve disease (AVD) using photoplethysmography (PPG) signals. This approach leverages approximately 170,000 unlabeled PPG recordings from the UK Biobank to construct pseudo-labels based on waveform phenotypes associated with aortic stenosis and regurgitation. After fine-tuning on a small labeled dataset, the model achieved AUROCs of 0.8025 for aortic stenosis and 0.7669 for aortic regurgitation, demonstrating its potential for low-cost, scalable AVD screening. AI

IMPACT This research demonstrates a new method for leveraging unlabeled physiological data to improve low-cost screening for cardiovascular conditions.

RANK_REASON The cluster contains an academic paper detailing a new machine learning method for a medical application. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• aortic valve disease
• aortic valve insufficiency
• Jiaze Wang
• photoplethysmogram
• Physiology-Guided Self-Supervised Learning
• UK Biobank

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →