PaperarXiv:2410.08559
ECG-JEPA: Learning General Representation of 12-Lead ECG with a Joint-Embedding Predictive Architecture
Neural Networks (Elsevier), published online 2026-07-07 (OpenAlex); arXiv v1 2024-10-11, v5 2026-04-10
Open codeOpen weights
(opens in a new tab)Kim, Sehun, "Learning General Representation of 12-Lead Electrocardiogram with a Joint-Embedding Predictive Architecture", arXiv:2410.08559, CC BY 4.0
TL;DR
JEPA for 12-lead ECG: a student encoder sees masked ECG, and a predictor regresses the teacher's latent features of masked patches. There is no raw-signal reconstruction. Cross-Pattern Attention (CroPA): a masked-attention prior for 12-lead structure. ViT-B-size encoders (12 layers, 16 heads, 768-d); predictor 6 layers, 384-d. Pretraining: Chapman + Ningbo (43,240 10-s ECGs, 500 Hz) + CODE-15 (130,900 10-s ECGs) ≈ 174k (paper: "≈180,000"). §9 ablation adds MIMIC-IV-ECG (960k).
- Form factor: clinical
Abstract
Electrocardiogram (ECG) captures the heart's electrical signals, offering valuable information for diagnosing cardiac conditions. However, the scarcity of labeled data makes it challenging to fully leverage supervised learning in the medical domain. Self-supervised learning (SSL) offers a promising solution, enabling models to learn from unlabeled data and uncover meaningful patterns. In this paper, we show that masked modeling in the latent space can be a powerful alternative to existing self-supervised methods in the ECG domain. We introduce ECG-JEPA, an SSL model for 12-lead ECG analysis that learns semantic representations of ECG data by predicting in the hidden latent space, bypassing the need to reconstruct raw signals. This approach offers several advantages in the ECG domain: (1) it avoids producing unnecessary details, such as noise, which is common in ECG; and (2) it addresses the limitations of naive L2 loss between raw signals. Another key contribution is the introduction of Cross-Pattern Attention (CroPA), a specialized masked attention mechanism tailored for 12-lead ECG data. ECG-JEPA is trained on the union of several open ECG datasets, totaling approximately 180,000 samples, and achieves state-of-the-art performance in various downstream tasks including diagnostic classification, feature extraction, and segmentation. Our code is openly available at https://github.com/sehunfromdaegu/ECG_JEPA.
Verbatim from arXiv (opens in a new tab) · arXiv metadata (abstracts are released under CC0)
Results on ChatPPG leaderboards
All leaderboardsNo results from this paper on ChatPPG leaderboards yet. Browse the leaderboards.
Tasks & signals
Signals
Keywords
Datasets
MIMIC-IV-ECGDiagnostic Electrocardiogram Matched Subset (v1.0)160,000 subjects · ECG
Chapman–Shaoxing–Ningbo 12-lead ECG arrhythmia database (PhysioNet)45,152 subjects · ECG- CODE-15%: large annotated 12-lead ECG dataset (Telehealth Network of Minas Gerais)ECG
PTB-XL21,799 clinical 12-lead ECGs (PhysioNet)18,869 subjects · ECG
Models
How it works
Cite this paper
APA
ECG-JEPA: Learning General Representation of 12-Lead ECG with a Joint-Embedding Predictive Architecture. (2027). Neural Networks. https://doi.org/10.1016/j.neunet.2026.109334
BibTeX
@article{ecgjepa12leadecg2027ecgjepa,
title = {{ECG-JEPA: Learning General Representation of 12-Lead ECG with a Joint-Embedding Predictive Architecture}},
year = {2027},
journal = {Neural Networks},
doi = {10.1016/j.neunet.2026.109334},
eprint = {2410.08559},
archivePrefix = {arXiv}
}Built from the hub record. Check it against the publisher before submitting.
Added to the ChatPPG research hub
