Key result
Self-supervised cardiac phase learning matches supervised U-Net precision for QRS delineation with ~10 ms error.
Why the study?
Accurate automated ECG delineation is challenging due to morphological variability, low amplitudes near wave boundaries, and onset/offset ambiguity, while most existing machine learning approaches rely on supervised learning requiring detailed expert annotations.
Does self-supervised cardiac cycle phase learning improve annotation-efficient automated ECG delineation of P, QRS, and T wave boundaries compared to classical and supervised delineation methods in ECG recordings?
Does self-supervised cardiac cycle phase learning improve annotation-efficient automated ECG delineation of P, QRS, and T wave boundaries compared to classical and supervised delineation methods in ECG recordings?
Self-supervised learning of the physiological cardiac cycle phase from unlabeled ECGs enables accurate localization of QRS boundaries using only 10 annotated calibration records and improves supervised ECG delineation when used for weight initialization.
May reduce annotation burden for ECG delineation; leaves open prospective validation before clinical use.
Computational methods play an important role in cardiology, supporting the analysis and interpretation of electrocardiographic data.Automated ECG delineation, including the localization of P-wave, QRS-complex, and T-wave boundaries, is particularly important for extracting clinically relevant temporal and morphological information from ECG recordings.However, accurate delineation remains challenging due to substantial variability in ECG morphology, low signal amplitudes near wave boundaries, and ambiguity in the definition of onset and offset points.Recent advances in machine learning show promising progress in biomedical data processing.Most existing machine learning approaches to ECG delineation rely on supervised learning and therefore require detailed expert annotations.In contrast, self-supervised methods can exploit large collections of unlabeled recordings.In this work, a self-supervised method for learning the physiological phase of the cardiac cycle directly from unlabeled ECG recordings is proposed.The model is trained to predict monotonically increasing phase positions on a unit circle, minimizing morphological distance for signals at similar phase points.Morphology is compared using delay embeddings defined at fixed phase lags.Monotonicity constraints further regularize physiologically consistent phase progression.The method is trained on 12-lead MIMIC-IV-ECG recordings without annotations and evaluated for ECG delineation on the QT Database.Without retraining, phase anchors calibrated on only 10 annotated records provide QRS-boundary localization with precision comparable to supervised models.The weights learned on the self-supervised task also improve a supervised delineation model when used for initialization.These results demonstrate the potential of self-supervised physiological phase learning as an annotation-efficient method for ECG delineation and further automated cardiac signal analysis.
No takes yet. Share an insight, caveat, or question.
A 2026 study studied Automated electrocardiogram (ECG) waveform delineation (n=104). Self-supervised cardiac cycle phase learning model (1D U-Net) vs. Supervised U-Net models and classical delineators (NeuroKit2 discrete wavelet transform and peak-prominence) was evaluated on QRS onset localization error standard deviation (ms) on the QT Database using raw phase crossings calibrated on 4 folds. Self-supervised cardiac phase learning localized QRS onset and offset with error SDs of 10.2 ms and 13.2 ms and 100% sensitivity, matching supervised U-Net precision (10.4 ms and 14.4 ms).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: