A Generalisable Heartbeat Classifier Leveraging Self-Supervised Learning for ECG Analysis During Magnetic Resonance Imaging

• Published in: IEEE journal of biomedical and health informatics 2024-09, Vol.28 (9), p.5147-5155
• Main Authors: Aublin, Pierre Gabriel, Felblinger, Jacques, Oster, Julien
• Format: Article
• Language: English
• Subjects: Adult; Algorithms; Deep Learning; Electrocardiogram; Electrocardiography; Electrocardiography - methods; Female; Heart beat; Heart Rate - physiology; Humans; magnethohydrodynamic effect; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Magnetohydrodynamics; Male; Middle Aged; Noise; Recording; representation learning; Siamese network; Signal Processing, Computer-Assisted; Supervised Machine Learning; Young Adult
• ISSN: 2168-2194; 2168-2208; 2168-2208
• DOI: 10.1109/JBHI.2024.3411792

Electrocardiogram (ECG) is acquired during Magnetic Resonance Imaging (MRI) to monitor patients and synchronize image acquisition with the heart motion. ECG signals are highly distorted during MRI due to the complex electromagnetic environment. Automated ECG analysis is therefore complicated in this context and there is no reference technique in MRI to classify pathological heartbeats. Imaging arrhythmic patients is hence difficult in MRI. Deep Learning based heartbeat classifier have been suggested but require large databases whereas existing annotated sets of ECG in MRI are very small. We proposed a Siamese network to leverage a large database of unannotated ECGs outside MRI. This was used to develop an efficient representation of ECG signals, further used to develop a heartbeat classifier. We extensively tested several data augmentations and self-supervised learning (SSL) techniques and assessed the generalization of the obtained classifier to ECG signals acquired in MRI. These augmentations included random noises and a model simulating MRI specific artefacts. SSL pretraining improved the generalizability of heartbeat classifiers in MRI (F1=0.75) compared to Deep Learning not relying on SSL (F1=0.46) and another classical machine learning approach (F1=0.40). These promising results seem to indicate that the use of SSL techniques can learn efficient ECG signal representation, and are useful for the development of Deep Learning models even when only scarce annotated medical data are available.