Noise-Robust Contrastive Learning with an MFCC-Conformer For Coronary Artery Disease Detection

2026-01-26

AI Summary

This study addresses the significant performance degradation of phonocardiogram (PCG)-based coronary artery disease (CAD) detection in real-world noisy environments. To mitigate this issue, the authors propose a multi-channel energy-driven approach for removing noise-contaminated segments, which uniquely integrates auxiliary noise-reference microphones with an energy-thresholding strategy during PCG preprocessing. Furthermore, they design an MFCC-Conformer deep classifier that leverages the Conformer architecture to effectively model multi-channel mel-frequency cepstral coefficient (MFCC) features, thereby enhancing noise robustness. By incorporating contrastive learning, the proposed method achieves 78.4% accuracy and 78.2% balanced accuracy on a dataset of 297 subjects, representing improvements of 4.1% and 4.3%, respectively, over the non-denoised baseline. This advancement substantially improves the robustness of CAD detection under non-stationary noise conditions.

Technology Category

• Machine Learning: Hardware-aware ML
• Computer Vision: Adversarial Attacks & Robustness
• Natural Language Processing: Safety and Robustness

Application Category

• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Large-scale security measurements

Abstract

Cardiovascular diseases (CVD) are the leading cause of death worldwide, with coronary artery disease (CAD) comprising the largest subcategory of CVDs. Recently, there has been increased focus on detecting CAD using phonocardiogram (PCG) signals, with high success in clinical environments with low noise and optimal sensor placement. Multichannel techniques have been found to be more robust to noise; however, achieving robust performance on real-world data remains a challenge. This work utilises a novel multichannel energy-based noisy-segment rejection algorithm, using heart and noise-reference microphones, to discard audio segments with large amounts of nonstationary noise before training a deep learning classifier. This conformer-based classifier takes mel-frequency cepstral coefficients (MFCCs) from multiple channels, further helping improve the model's noise robustness. The proposed method achieved 78.4% accuracy and 78.2% balanced accuracy on 297 subjects, representing improvements of 4.1% and 4.3%, respectively, compared to training without noisy-segment rejection.

Problem

• coronary artery disease
• phonocardiogram
• noise robustness
• multichannel
• nonstationary noise

Innovation

• noise-robust contrastive learning
• MFCC-Conformer
• multichannel noisy-segment rejection
• phonocardiogram
• coronary artery disease detection