ECG-Expert-QA

该数据集是一个专为评估心电图解释中诊断能力而设计的一体化多模态数据集，它融合了真实的临床数据与系统生成的合成案例。该数据集包含了47,211个精心筛选的问题-答案对。它覆盖了从基础心律分析到复杂病例解释等多种临床场景，为医学问答和智能诊断系统提供了高质量的训练数据。其规模达到47,211个问题-答案对，任务旨在评估在心电图解释中诊断准确性、临床推理和知识整合的能力。This integrated multimodal dataset is specifically designed to evaluate diagnostic capabilities in electrocardiogram (ECG) interpretation by combining real clinical data and system-generated synthetic cases. It contains 47,211 carefully curated question-answer pairs, covering a wide range of clinical scenarios ranging from basic cardiac rhythm analysis to complex case interpretation, providing high-quality training data for medical question answering and intelligent diagnostic systems. With a total of 47,211 question-answer pairs, this dataset supports tasks that aim to assess diagnostic accuracy, clinical reasoning, and knowledge integration abilities in ECG interpretation.

搜集汇总

背景与挑战

背景概述

ECG-Expert-QA是一个用于评估医学大型语言模型在心电图领域表现的基准数据集，包含多个方面的医学问答和诊断任务，与MIMIC-IV-ECG数据集相关，并提供了预处理代码和使用说明。