计算机辅助医学 医学知识库 李作峰 上海生物信息技术研究中心

提纲 基本概念 1 关键技术 2 应用实例 3

基本概念  医学知识的来源 科学知识 - 认知和推理等 经验知识 - 专家临床经验等 医生根据具体病人的症状和体征进行推理的时候， 这两种类知识是相互交织的。 科学知识 经验知识 临床诊疗

经验知识 科学知 识 传统 诊疗 传统医学以医生的经验为主导 误诊 漏诊 晚诊 不合理用药 问题 基本概念

医学知识爆炸 医生个体医学知识与临 床循证知识的巨大鸿沟 医学知识数量 医生个体医学知识 基本概念

 知识库的定义 A knowledge base is a special kind of database for knowledge management, providing the means for the computerized collection, organization, and retrieval of knowledge.

基本概念  知识获取的方式 知识获取 知识获取 (knowledge acquisition) 是指从某个领域或某些知 识源 ( 如专家、书本或专家处理的实例 ) 获取专家系统实现 问题求解所需要的专门知识，包括共性知识和个性知识。 抽提、归纳、总结 知识库  直接获取  间接获取

基本概念  知识获取的步骤 确定领域问题 概念化 形式化和编码 测试 再构造 再设计 再构造

知识库构建的关键问题  知识的完整性  知识的安全性  知识获取工具  知识管理

医学知识库构建的关键技术  医学知识模型的构建 适合医学知识表达、存储、检索和共享的医学 知识模型 知识模型应涵盖医学知识实体及实体间的关系 本体 （ ontology ）

医学知识库构建的关键技术  An ontology is an explicit description of a domain: concepts properties and attributes of concepts constraints on properties and attributes Individuals (often, but not always)  An ontology defines a common vocabulary a shared understanding 本体（ ontology ）

医学知识库构建的关键技术  To share common understanding of the structure of information among people among software agents  To enable reuse of domain knowledge to avoid “re-inventing the wheel” to introduce standards to allow interoperability Why Develop an Ontology? 本体（ ontology ）

医学知识库构建的关键技术  To make domain assumptions explicit easier to change domain assumptions (consider a genetics knowledge base) easier to understand and update legacy data  To separate domain knowledge from the operational knowledge re-use domain and operational knowledge separately (e.g., configuration based on constraints) Why Develop an Ontology? 本体（ ontology ）

Ontologies Software agents Problem- solving methods Domain- independent applications Domain- independent applications Databases Declare structure Knowledge bases Knowledge bases Provide domain description 医学知识库构建的关键技术 An Ontology Is Often Just the Beginning 本体（ ontology ）

医学知识库构建的关键技术 本体（ ontology ） Biomedical ontology in action  Knowledge Management Annotating Data and Resources  Biomedical ontology serves as a source of vocabulary for the purpose of annotating data or indexing documents.  Indexing is principally used in reference to the assignment of entries from a controlled vocabulary to documents  The indexing of clinical documents is generally referred to as coding  In biology, the functional description of experimental data is usually referred to as annotation Accessing Biomedical Information Mapping across Biomedical Ontologies

医学知识库构建的关键技术 本体（ ontology ） Biomedical ontology in action  Knowledge Management Annotating Data and Resources Accessing Biomedical Information The main function of the indexing of large document collections such as MEDLINE is to support accurate retrieval Several biomedical search engines exploit MeSH and the UMLS to provide access to the biomedical literature The automatic classification of biomedical documents is also generally supported by ontologies. Mapping across Biomedical Ontologies

医学知识库构建的关键技术 本体（ ontology ） Biomedical ontology in action  Knowledge Management Annotating Data and Resources Accessing Biomedical Information Mapping across Biomedical Ontologies  users can choose from a variety of ontologies and select the artifact that best fits their purpose.  resources annotated to different ontologies become more difficult to integrate, unless mappings are created among ontologies in order to identify equivalent concepts across ontologies.

医学知识库构建的关键技术 本体（ ontology ） Biomedical ontology in action  Data Integration, Exchange and Semantic Interoperability Information Exchange and Semantic Interoperability  RxNorm, UMLS, SNOMED CT clinical information systems  LONIC exchange of laboratory data  The HL7 Clinical Document Architecture, Release 2 (CDA R2) model is "richly expressive, enabling the formal representation of clinical statements" Information and Data Integration

医学知识库构建的关键技术 本体（ ontology ） Biomedical ontology in action  Data Integration, Exchange and Semantic Interoperability Information Exchange and Semantic Interoperability Information and Data Integration  ontology support the standardization required from warehousing approaches to data integration  mediation-based approaches use ontologies for defining a global schema and mapping between the global schema and local schemas  ontologies facilitate the integration of datasets, often by providing a common reference for biomedical entities in several datasets.

医学知识库构建的关键技术 本体（ ontology ） Biomedical ontology in action  Decision Support and Reasoning Data Selection By providing an abstraction of some domain, ontologies can help define groups from a high level value for the independent variable, instead of listing all possible values Data Aggregation Decision Support Natural Language Processing Applications Knowledge Discovery breast cancer cancer of upper-inner quadrant of breast cancer of lower-outer quadrant of breast

医学知识库构建的关键技术 本体（ ontology ） Biomedical ontology in action  Decision Support and Reasoning Data Selection Data Aggregation  ontologies are used for identifying the characteristics of groups obtained through various methods  ICD 10 Decision Support Natural Language Processing Applications Knowledge Discovery

医学知识库构建的关键技术 本体（ ontology ） Biomedical ontology in action  Decision Support and Reasoning Data Selection Data Aggregation Decision Support  ontologies provide a standard vocabulary for biomedical entities, helping standardize and integrate data sources  ontologies are a source of computable domain knowledge that can be exploited for decision support purposes, often in combination with business rules Natural Language Processing Applications Knowledge Discovery

医学知识库构建的关键技术 本体（ ontology ） Biomedical ontology in action  Decision Support and Reasoning Data Selection Data Aggregation Decision Support Natural Language Processing Applications Knowledge Discovery NLPontology term recognition domain knowledge

医学知识库构建的关键技术 本体（ ontology ） Biomedical ontology in action  Decision Support and Reasoning Data Selection Data Aggregation Decision Support Natural Language Processing Applications Knowledge Discovery ontologies are a component of the data-driven approach to biomedical research ontologies have been used for identifying relations between genotype and phenotype, Data mining

医学知识库构建的关键技术 文本知识发现 (Knowledge Discovery in Texts) 就是从文本 集中发现和挖掘归纳性的知识, 如有用的模式、模型、 趋势、规则等知识 生物医学 文本 电子病历 健康档案 医学文献 …… NLP Machine Learning 知识库知识库 文本挖掘

医学知识库构建的关键技术  生物医学文献挖掘这个概念最早在 1986 年由芝加 哥大学教授 Don R.Swan- son 提出 文本挖掘 A A B B C C B B C C A A 镁缺乏 偏头疼

医学知识库构建的关键技术 文本挖掘 评测标准 召回率 准确率 F- 测度 一般流程  信息检索 （ Information Retrieval ， IR ）  实体识别 （ Entity Recognition ， ER ）  信息抽提 （ Information Extraction ， IE ）

医学知识库构建的关键技术  知识的存储  知识的查询  知识的展示  知识的更新 知识库构建

医学知识库  MD Consult-Medical Doctor Consult  该数据库被称为 “ 临床医学知识库 ” 、 “ 内科 医师的在线临床工具 ” ，是临床卫生科技人 员的循证医学资源参考数据库。  MD Consult 由世界最大的三家英文医学出版 商 Mosby 、 Lippincott Williams ＆ Wilkins 、 W ． B ． Saunders 联合创建，由世界上五大 出版社之一的 Elsevier Science 出版发行，于 1997 年开始正式投入使用。

 MD Consult 是一个集成医学文献、图书、临床指 南、药物等的综合性医学知识库 一、 First Consult 二、参考书籍（ Books ） 三、医学期刊（ Journals ） 四、北美临床杂志（ The Clinics ） 五、病人教育（ Patient Education ） 六、药物库（ Drugs ） 七、诊疗指南（ Guidelines ） 八、医学图片（ Images ） 九、医学新闻（ News ） 十、继续医学教育（ CME ） 医学知识库

 常见症状列表

 MD Consult 是一个集成医学文献、图书、临床指 南、药物等的综合性医学知识库 Keywords ： colorectal cancer 医学知识库

 问答系统 医学知识库

 问答系统 医学知识库 NLP Vocabulary

 基于知识库的临床决策支持系统 医学知识库 规则库 条件库 动态规则库

医学知识库  Mayo Clinic-Health information Disease and condition Symptoms Drugs and supplements Tests and procedures Healthy lifestyle

医学知识库

 Symptom checker

医学知识库  Drugs and supplements

医学知识库  Drugs

医学知识库  Tests

中文医学知识库  翻译 CD003255/abstract

中文医学知识库  翻译 CD003255/abstract

中文医学知识库  大型综合型医院网站 举例：瑞金医院 1. 药物信息：  x x 2. 诊疗常规  医学中心  各种常见病介绍  医学中心 医学中心 

课程作业讨论

参考资料  蒋立辉, 王 伟. 医学 知 识 库 与 医学知识的 获取.  O. Bodenreider. Biomedical Ontologies in Action: Role in Knowledge Management, Data Integration and Decision Support.  Daniel L. Rubin, Nigam H. Shah and Natalya F. Noy. Biomedical ontologies: a functional perspective.

谢谢！