计算机工程与应用 ›› 2023, Vol. 59 ›› Issue (17): 35-47.DOI: 10.3778/j.issn.1002-8331.2212-0285

• 热点与综述 • 上一篇    下一篇

中医症状信息抽取研究综述

易钧汇,查青林   

  1. 江西中医药大学 计算机学院,南昌 330004
  • 出版日期:2023-09-01 发布日期:2023-09-01

Survey of TCM Symptom Information Extraction

YI Junhui, ZHA Qinglin   

  1. School of Computer Science, Jiangxi University of Chinese Medicine, Nanchang 330004, China
  • Online:2023-09-01 Published:2023-09-01

摘要: 随着我国医疗信息化水平的提高以及电子病历普及,医疗数据量呈爆炸式增长的趋势。中医电子病历是中医医案的结构化存储形式,其中包含大量的中医临床经验。症状是医家诊病、辨证的主要依据。对中医电子病历进行症状信息抽取可以得到文本中的关键信息,对提高中医诊疗效率和后续病症、药症关系等研究分析工作提供帮助。简要介绍中医症状信息抽取流程;分别阐述中医症状命名实体识别和关系抽取的难点、评价标准和近年的研究成果,并对这些研究成果所采用的方法进行了对比分析;总结症状信息抽取的下游应用,给出症状信息抽取任务中问题的解决思路。

关键词: 中医, 信息抽取, 命名实体识别, 关系抽取, 症状

Abstract: With the improvement of China’s medical information and the popularity of electronic medical records, the volume of medical data is on an explosive growth trend. TCM electronic medical record is a structured storage form of the TCM cases, which contains a lot of clinical experiences. The doctors diagnose diseases according to the symptoms. It can obtain the key information of the text to extract symptom information from TCM electronic medical records, which can offer help for improving the efficiency of TCM treatment and the relative research on the herbs-symptoms and diseases-symptoms relationship. Firstly, the process of TCM symptom information extraction is briefly introduced in this paper. Then, the primary problems that named entity recognition and relation extraction of TCM symptom, evaluation criteria and research results in recent years are respectively analyzed, and a comparative analysis is used on the methods of the research results. Finally, the downstream applications of symptom information extraction are summarized, and ideas for solving problems in symptom information extraction tasks are offered.

Key words: traditional Chinese medicine(TCM), information extraction, named entity recognition, relation extraction, symptom