Holtercare-Bench：评估长期动态心电图分析的多模态基准
原标题：Holtercare-Bench: A Multimodal Benchmark for Evaluating Long-Term Dynamic ECG Analysis
AI 摘要
浙江大学研究团队提出了Holtercare-Bench，一个用于评估长期动态心电图分析的多模态基准，并配套发布了包含788份临床Holter记录和22,980个问答对的Holtercare-23K数据集。该基准通过信号-视频-文本三模态对齐，评估模型在时间定位、临床诊断和全局总结方面的能力。零样本评估显示主流多模态大模型在处理超长病理序列时存在显著性能差距，而微调后性能大幅提升。该工作揭示了当前多模态大模型在心电生理学领域的局限性，为长期医疗多模态大模型提供了基础基准。
正文节选
Holtercare-Bench: A Multimodal Benchmark for Evaluating Long-Term Dynamic ECG Analysis Abstract. ††footnotetext: ∗These authors contributed equally to this research. ††footnotetext: †Corresponding authors. While multimodal large language models (MLLMs) excel in medical applications, most of them favor static images or short-term signals. In the critical field of dynamic electrocardiograms (ECG), models struggle with complex temporal reasoning and diagnostic report generation due to a lack of hig