PMC-Patients
收藏资源简介:
PMC-Patients是由清华大学开发的大型患者摘要和关系数据集,包含从PubMed Central提取的167,000份患者摘要,旨在为基于检索的临床决策支持系统提供基准测试。数据集通过简单的启发式方法从案例报告中提取患者摘要,并利用PubMed引用图定义患者文章相关性和患者间相似性。PMC-Patients不仅规模庞大,而且覆盖了广泛的医疗条件,适用于评估患者到文章检索和患者到患者检索任务,展示了在临床决策支持中的实际应用价值。
PMC-Patients is a large-scale patient summary and relational dataset developed by Tsinghua University. It contains 167,000 patient summaries extracted from PubMed Central, and is designed to provide benchmarks for retrieval-based clinical decision support systems. The dataset extracts patient summaries from case reports via simple heuristic methods, and leverages PubMed citation graphs to define patient-article relevance and inter-patient similarity. With its large scale and wide coverage of diverse medical conditions, PMC-Patients is suitable for evaluating patient-to-article retrieval and patient-to-patient retrieval tasks, and demonstrates its practical application value in clinical decision support.




