基于NLP构建病历后结构化专病数据库探索与实践  

Exploration and Practice of Constructing a Structured Specialized Disease Database Based on NLP for Medical Records

作　　者：张亚男[1] 董亮[1] 何萍 ZHANG Yanan;DONG Liang;HE Ping(Longhua Hospital,Shanghai University of Traditional Chinese Medicine,Shanghai 200032,China;Shanghai Hospital Development Center,Shanghai 200041,China)

机构地区：[1]上海中医药大学附属龙华医院,上海200032 [2]上海申康医院发展中心,上海200041

出　　处：《医学信息学杂志》2024年第9期82-86,共5页Journal of Medical Informatics

基　　金：上海申康医院发展中心管理研究项目(项目编号:2023SKMR-22)。

摘　　要：目的/意义建设基于结构化电子病历的专病数据库,提高专病数据库规范性和可用性,提高临床科研工作效率。方法/过程采用模板化输入、自然语言处理等技术,将非结构化电子病历转化为结构化电子病历,基于结构化电子病历构建专病数据库。结果/结论龙华医院基于结构化电子病历建设的银屑病专病数据库分中心,为临床科研人员提供结构化科研数据源,辅助提升分析效率;同时有效支撑上海申康“基于多中心的银屑病专病大数据临床科研随访一体化平台”建设,有助于专病数据库高质量、规模化发展。Purpose/Significance To construct a specialized disease database based on structured electronic medical records(EMR),and to improve the standardization and usability of the specialized disease database and improve the efficiency of clinical scientific research.Method/Process By using templated input,natural language processing(NLP)and other technologies,unstructured EMR are converted into structured EMR,and a specialized disease database is constructed based on structured EMR.Result/Conclusion The psoriasis specialized disease database sub center built by Longhua Hospital based on structured EMR provides clinical researchers with a structured scientific research data source,assists researchers in improving analysis efficiency,and effectively supports the construction of Shanghai Shenkang’s“multi center psoriasis specialized disease big data clinical research follow-up integrated platform”,which is conducive to the high-quality and large-scale development of specialized disease databases.

关 键 词：自然语言处理 结构化电子病历 专病数据库 

分 类 号：R-058[医药卫生]