Foundation Model for Cardiac Time Series via Masked Latent Attention
通过掩码潜在注意力的心脏时间序列基础模型
机构 * Department of Computer Science, ETH Zurich, Switzerland(苏黎世联邦理工学院计算机科学系,瑞士) ; Department of Biomedical Engineering, University of Basel, Switzerland(巴塞尔大学生物医学工程系,瑞士)
专题命中 健康监测 :diagnosis(abstract);分类 cs.LG、eess.SP
AI总结 本文提出LAMAE基础模型,通过自监督预训练学习跨导联连接机制,提升ECG表示质量和迁移能力,在预测ICD-10编码上优于独立导联掩码模型和对齐基线。
Comments First two authors are co-first. Last two authors are co-senior
Journal ref ICLR 2026: Workshop on Foundation Models for Science