Domain Knowledge Is Power: Leveraging Physiological Priors for Self-Supervised Representation Learning in Electrocardiography
IEEE Transactions on Biomedical Engineering · 2026
DOI: 10.1109/tbme.2026.3656904 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Biomedical Engineering, 2026. Verfasst von Nooshin Maghsoodi und 10 weiteren Autor:innen. Hauptthemen laut MEDLINE: Electrocardiography, Supervised Machine Learning, Signal Processing, Computer-Assisted. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 48 Literaturangaben. Gefördert von: Social Sciences and Humanities Research Council, Canadian Institutes of Health Research, Southeastern Ontario Academic Medical Association.
Abstract (Originalsprache)
OBJECTIVE: Electrocardiograms (ECGs) play a crucial role in diagnosing heart conditions; however, the effectiveness of artificial intelligence (AI)-based ECG analysis is often hindered by the limited availability of labeled data.Self-supervised learning (SSL) can address this by leveraging large-scale unlabeled data. We introduce PhysioCLR (Physiology-aware Contrastive Learning Representation for ECG), a physiology-aware contrastive learning framework that incorporates domain-specific priors to enhance the generalizability and clinical relevance of ECG-based arrhythmia classification. METHODS: During pretraining, PhysioCLR learns to bring together embeddings of samples that share similar clinically relevant features while pushing apart those that are dissimilar. Unlike existing methods, our method integrates ECG physiological similarity cues into contrastive learning, promoting the learning of clinically meaningful representations. Additionally, we introduce ECG-specific augmentations that preserve the ECG category post-augmentation and propose a hybrid loss function to further refine the quality of learned representations. RESULTS: We evaluate PhysioCLR on two public ECG datasets, Chapman and Georgia, for multilabel ECG diagnoses, as well as a private ICU dataset labeled for binary classification. Across the Chapman, Georgia, and private cohorts, PhysioCLR boosts the mean AUROC by 12% relative to the strongest baseline, underscoring its robust cross-dataset generalization. CONCLUSION: By embedding physiological knowledge into contrastive learning, PhysioCLR enables the model to learn clinically meaningful and transferable ECG features. SIGNIFICANCE: PhysioCLR demonstrates the potential of physiology-informed SSL to offer a promising path toward more effective and label-efficient ECG diagnostics.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Förderung
- Social Sciences and Humanities Research Council
- Canadian Institutes of Health Research
- Southeastern Ontario Academic Medical Association
- Canada CIFAR AI Chair
- Canada Research Chair
- Natural Sciences and Engineering Research Council of Canada
- New Frontiers in Research Fund
- Vector Institute
Zitieren
APA
Maghsoodi, N., Nassar, S., Wilson, P. F. R., To, M. N. N., Mannina, S., Addas, S., Sibley, S., Pichora, D., Maslove, D., Abolmaesumi, P., & Mousavi, P. (2026). Domain Knowledge Is Power: Leveraging Physiological Priors for Self-Supervised Representation Learning in Electrocardiography. IEEE Transactions on Biomedical Engineering. https://doi.org/10.1109/tbme.2026.3656904
Vancouver
Maghsoodi N, Nassar S, Wilson PFR, To MNN, Mannina S, Addas S, et al. Domain Knowledge Is Power: Leveraging Physiological Priors for Self-Supervised Representation Learning in Electrocardiography. IEEE Transactions on Biomedical Engineering. 2026. doi:10.1109/tbme.2026.3656904
BibTeX
@article{Maghsoodi2026,
author = {Maghsoodi, Nooshin and Nassar, Sarah and Wilson, Paul F R and To, Minh Nguyen Nhat and Mannina, Sophia and Addas, Shamel and Sibley, Stephanie and Pichora, David and Maslove, David and Abolmaesumi, Purang and Mousavi, Parvin},
title = {Domain Knowledge Is Power: Leveraging Physiological Priors for Self-Supervised Representation Learning in Electrocardiography},
journal = {IEEE Transactions on Biomedical Engineering},
year = {2026},
doi = {10.1109/tbme.2026.3656904}
}
EUVIMED