LAMAE：面向多模态心脏表征学习的潜在注意力掩码自编码器
原标题：Reading the Whole Heart: Latent-Attention Masked Autoencoders for Multimodal Cardiac Representation Learning
AI 摘要
研究者提出 Latent-Attention Masked Autoencoders（LAMAE），一种多模态、结构感知的掩码自编码器，在自监督预训练阶段通过共享的潜在注意力模块在 study–view–entity 层级上直接交换信息，联合学习患者级表征。该模型在超过 120 万条 MIMIC-IV 住院记录上预训练，在多模态住院任务（院内死亡率、ICD-10 与 DRG 编码、住院时长）上优于模态特定预训练及强对比学习和视觉-语言基线，且在测试时仅有一种模态可用时仍保持优势。
正文节选
Reading the Whole Heart: Latent-Attention Masked Autoencoders for Multimodal Cardiac Representation Learning Abstract Cardiovascular diagnosis rests on integrating complementary modalities, like ECG, echocardiography, chest radiographs, and clinical variables, each capturing distinct but correlated aspects of cardiac physiology. Yet most medical foundation models remain modality-specific, combining modalities only for finetuning or post-training. This discards the cross-modal evidence clinician