Semantic indexing of biomedical publications and clinical text

生物医学出版物和临床文本的语义索引

• 批准号: 10269685
• 负责人: Dina Demner-Fushman
• 金额: $ 59.94万
• 依托单位: NATIONAL LIBRARY OF MEDICINE
• 依托单位国家: 美国
• 资助国家: 美国
• 起止时间: 至
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10269685
• 关键词: Address; Algorithms; Anatomy; Animal Model; Automated Indexing; Chemicals; Clinical; Collaborations; Disease; Elements; Location; Manuals; MeSH Thesaurus; Medical; Methods; Names; Pharmaceutical Preparations; Publications; Research; Research Personnel; Role; Schedule; Semantics; Services; Study Subject; Suggestion; Text; Triage; improved; indexing; language processing; programs; tool

The Indexing Initiative project addresses both objectives by continually improving Named Entity Recognition algorithms and providing these algorithms to the public in a program called MetaMap. A recently developed MetaMap Lite, a faster and more accurate tool for Named Entity Recognition, continues gaining popularity and is gradually replacing the original MetaMap, which is schedules for sunsetting in FY 2021-22. Both MetaMap and MetaMap Lite are widely used as NLM services for remote file processing and as downloadable tools. The semantic indexing objective is addressed by the Medical Text Indexer (MTI) program that assigns most salient medical subject headings to medical text. Among other applications, MTI powers MeSH on Demand an online tool that helps authors find key terms for their publications, identifying MeSH Terms relevant to submitted text. In FY 2020, in collaboration with the Index section, and NCBI, LHC researchers have developed and deployed a tool that automatically excludes out-of-scope articles from the Indexers queue. The reduction in manual triage of these articles is close to 50%, which amounts to about 40,000 articles. In FY 2020 LHC researchers significantly improved chemical name recognition in biomedical publications and improved suggestions for qualifiers of medical subject headings by about 30%.

索引计划项目通过不断改进命名的实体识别算法并在称为MetAmap的程序中向公众提供这些算法来解决这两个目标。最近开发的MetAmap Lite是一种更快，更准确的命名实体识别工具，它继续越来越受欢迎，并正在逐渐取代原始的MetAmap，该MetAmap是在2021 - 22财年中日落的时间表。 MetAmap和MetAmap Lite均被广泛用作远程文件处理和可下载工具的NLM服务。语义索引目标由医学文本索引器（MTI）计划解决，该计划将最显着的医学主题标题分配给医学文本。在其他应用程序中，MTI为网格按需提供网络工具，该工具可帮助作者找到其出版物的关键术语，并确定与已提交的文本相关的网格术语。 在2020财年，与索引部分合作，NCBI和LHC研究人员开发并部署了一种工具，该工具自动将索引者排队排除在索引队列中。这些文章的手动分类减少接近50％，约为40,000篇文章。 在2020财年，LHC研究人员在生物医学出版物中显着提高了化学名称识别，并改善了医学主题标题预选赛的建议，大约30％。