Clinical and Informatics Research in Medical Terminologies

医学术语的临床和信息学研究

• 批准号: 10268077
• 负责人: Kin Wah Fung
• 金额: $ 61.02万
• 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/10268077
• 关键词: Adoption; Algorithms; Area; Certification; Classification; Clinical Informatics; Clinical Research; Code; Collaborations; Common Data Element; Communication; Computing Methodologies; Data; Data Reporting; Data Science; Data Sources; Diagnosis; Drug Interactions; Drug Labeling; Electronic Health Record; Graph; HIV; Health; Health Personnel; ICD-10-CM; Information Retrieval; International; International Classification of Disease Codes; International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10); Intervention; Knowledge; Label; Machine Learning; Maps; Measurement; Medical; Medical Research; Methods; Natural Language Processing; Ontology; Output; Patient Care; Patients; Pattern; Pharmaceutical Preparations; Procedures; Publications; Publishing; Research; Resources; Retrieval; SNOMED Clinical Terms; Semantics; Source; Standardization; Structure; System; Techniques; Terminology; Text; big-data science; care outcomes; clinical decision support; data integration; data reuse; data sharing; data standards; data structure; deep learning; improved; incentive program; indexing; information organization; insight; interoperability; lexical; statistics; structured data; symposium; tool

A. Extraction of information from drug labels using natural language processing Drug package inserts (drug labels) are a comprehensive, up-to-date and authoritative source of drug information that is publicly available. To unleash the knowledge in the drug labels, they need to be transformed into standardized data structure and encoded in standard terminologies. Only then can the knowledge be used to drive applications such as clinical decision support. This collaborative project with the FDA uses natural language processing (NLP) and machine learning to extract information from the drug labels and create mappings to standard terminologies. The output will support the FDAs drug label indexing initiative to increase the usefulness of drug labels. In collaboration with FDA, I have hosted a challenge for extracting drug-drug interaction information from drug labels for the second year through NISTs Text Analysis Conference (TAC2019). B. Use of medical terminologies to support clinical research Medical terminologies and common data elements are important tools to allow sharing of clinical research data. I have studied their application in data sharing involving HIV-infected patients. C. Creating maps between commonly used terminologies Mapping provides a solution to the problem caused by the use of multiple coding systems for the same kind of information. One example is the use of SNOMED CT and ICD-10-CM for coding medical diagnosis and problems. Using various computational methods supplemented by expert review, I have developed maps between SNOMED CT and the different flavors and versions of ICD codes. This will help to facilitate data re-use and data integration. I have also studied the potential benefits of using maps in data encoding. I am also studying various algorithmic approaches to create mappings between SNOMED CT and ICD-10-PCS, including lexical matching, ontological alignment and indirect mapping. On request from National Committee on Vital and Health Statistics, I have studied the new ICD-11 and compared it to ICD-10 and ICD-10-CM. An ongoing study is comparing the newly-released International Classification of Health Interventions (ICHI, created by WHO) with other medical procedure terminologies. D. Use of deep learning in terminology research Deep learning techniques have resulted in breakthroughs in many areas. I have started a new line of research to employ deep learning techniques (e.g., word and graph embeddings) to tasks such as semantic relatedness measurement and mapping. E. Facilitating adoption of terminology standards According to the Meaningful Use and subsequent Improving Interoperability incentive programs, SNOMED CT and RxNorm are terminologies required for the certification of electronic health record systems. I have studied the practical barriers of adoption of these terminologies and created useful resources to help with implementation. I studied the usage pattern of SNOMED CT terms in the problem lists of large health care providers and published a list of the most commonly used terms as the CORE Problem List Subset of SNOMED CT. The CORE subset is not only a useful resource for SNOMED CT implementers, it is also frequently used for terminology research and other purposes, and cited in multiple publications. RxTerms is another resource that I have developed to overcome data entry problems with RxNorm.

A. 使用自然语言处理从药品标签中提取信息 药品说明书（药品标签）是公开的全面、最新、权威的药品信息来源。为了释放药品标签中的知识，需要将它们转换为标准化数据结构并用标准术语进行编码。只有这样，这些知识才能用于驱动临床决策支持等应用。这个与 FDA 的合作项目使用自然语言处理 (NLP) 和机器学习从药物标签中提取信息并创建到标准术语的映射。该成果将支持 FDA 的药品标签索引计划，以提高药品标签的实用性。我与 FDA 合作，连续第二年通过 NIST 文本分析会议 (TAC2019) 主办了一项从药物标签中提取药物相互作用信息的挑战赛。 B. 使用医学术语支持临床研究 医学术语和通用数据元素是共享临床研究数据的重要工具。我研究了它们在涉及艾滋病毒感染者的数据共享中的应用。 C. 创建常用术语之间的映射 映射为同一类信息使用多种编码系统所带来的问题提供了解决方案。一个例子是使用 SNOMED CT 和 ICD-10-CM 来编码医疗诊断和问题。使用各种计算方法并辅以专家评审，我开发了 SNOMED CT 与不同风格和版本的 ICD 代码之间的地图。这将有助于促进数据重用和数据集成。我还研究了在数据编码中使用映射的潜在好处。我还在研究各种算法方法来创建 SNOMED CT 和 ICD-10-PCS 之间的映射，包括词汇匹配、本体对齐和间接映射。应国家生命与健康统计委员会的要求，我研究了新的 ICD-11，并将其与 ICD-10 和 ICD-10-CM 进行了比较。一项正在进行的研究正在将新发布的国际健康干预分类（ICHI，由世界卫生组织创建）与其他医疗程序术语进行比较。 D. 深度学习在术语研究中的应用 深度学习技术在许多领域取得了突破。我已经开始了一项新的研究，将深度学习技术（例如，单词和图形嵌入）应用于语义相关性测量和映射等任务。 E. 促进术语标准的采用 根据有意义的使用和随后的改进互操作性激励计划，SNOMED CT 和 RxNorm 是电子健康记录系统认证所需的术语。我研究了采用这些术语的实际障碍，并创建了有用的资源来帮助实施。我研究了大型医疗保健提供者的问题列表中 SNOMED CT 术语的使用模式，并发布了最常用术语的列表作为 SNOMED CT 的核心问题列表子集。 CORE 子集不仅是 SNOMED CT 实施者的有用资源，还经常用于术语研究和其他目的，并在多个出版物中引用。 RxTerms 是我开发的另一个资源，用于克服 RxNorm 的数据输入问题。