Foundation Model for Cardiac Time Series via Masked Latent Attention

arXiv cs.AI / 3/30/2026

💬 OpinionSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper proposes LAMAE, a foundation model for ECG time series that uses masked latent attention to explicitly capture cross-lead structural redundancy rather than treating leads as independent channels.
During self-supervised pretraining, LAMAE learns cross-lead connection mechanisms and supports higher-order cross-lead interactions, including permutation-invariant aggregation and adaptive weighting of lead-specific features.
Experiments on the MIMIC-IV-ECG dataset show that modeling cross-lead connections provides a form of structural supervision that improves learned representations and their transferability.
For downstream tasks, the model demonstrates strong performance in predicting ICD-10 codes and outperforms independent-lead masked modeling as well as alignment-based baselines.

Abstract

Electrocardiograms (ECGs) are among the most widely available clinical signals and play a central role in cardiovascular diagnosis. While recent foundation models (FMs) have shown promise for learning transferable ECG representations, most existing pretraining approaches treat leads as independent channels and fail to explicitly leverage their strong structural redundancy. We introduce the latent attention masked autoencoder (LAMAE) FM that directly exploits this structure by learning cross-lead connection mechanisms during self-supervised pretraining. Our approach models higher-order interactions across leads through latent attention, enabling permutation-invariant aggregation and adaptive weighting of lead-specific representations. We provide empirical evidence on the Mimic-IV-ECG database that leveraging the cross-lead connection constitutes an effective form of structural supervision, improving representation quality and transferability. Our method shows strong performance in predicting ICD-10 codes, outperforming independent-lead masked modeling and alignment-based baselines.