Search papers, labs, and topics across Lattice.
RhythmBERT, a self-supervised ECG language model, encodes ECG waveforms into symbolic tokens representing P, QRS, and T segments using an autoencoder, combined with continuous embeddings to preserve morphology. By pretraining on 800,000 unlabeled ECG recordings with a masked prediction objective, RhythmBERT learns contextual representations of ECG rhythms. The model achieves performance comparable to or exceeding strong 12-lead ECG baselines using only a single lead, demonstrating effective generalization across various heart conditions.
RhythmBERT treats ECGs as a language, achieving state-of-the-art heart disease detection with a single lead by encoding P, QRS, and T segments into symbolic tokens.
Electrocardiogram (ECG) analysis is crucial for diagnosing heart disease, but most self-supervised learning methods treat ECG as a generic time series, overlooking physiologic semantics and rhythm-level structure. Existing contrastive methods utilize augmentations that distort morphology, whereas generative approaches employ fixed-window segmentation, which misaligns cardiac cycles. To address these limitations, we propose RhythmBERT, a generative ECG language model that considers ECG as a language paradigm by encoding P, QRS, and T segments into symbolic tokens via autoencoder-based latent representations. These discrete tokens capture rhythm semantics, while complementary continuous embeddings retain fine-grained morphology, enabling a unified view of waveform structure and rhythm. RhythmBERT is pretrained on approximately 800,000 unlabeled ECG recordings with a masked prediction objective, allowing it to learn contextual representations in a label-efficient manner. Evaluations show that despite using only a single lead, RhythmBERT achieves comparable or superior performance to strong 12-lead baselines. This generalization extends from prevalent conditions such as atrial fibrillation to clinically challenging cases such as subtle ST-T abnormalities and myocardial infarction. Our results suggest that considering ECG as structured language offers a scalable and physiologically aligned pathway for advancing cardiac analysis.
