基于临床指南的高血压医学知识库设计研究

发布时间：2018-04-25 00:35

本文选题：临床指南 + 本体　；参考：《北京协和医学院》2017年硕士论文

【摘要】：研究目的将静态文本形式的临床指南按照合适的指南表达模型进行建模形成计算机可理解的形式,将本体理论和知识库技术结合起来,把文本形式临床指南转换成方便推理的知识库,有助于提高临床指南的使用效率,减少临床实践活动中医生的诊疗失误。以此为基础也可以作为高血压临床决策支持系统的重要模块,降低设计构建的难度。研究内容1.医学知识库构建相关理论对比分析通过对医学知识库构建的相关技术总结分析,找到适合高血压医学知识库设计与构建的模型与技术。如本体构建的语言和工具及本体推理机等,分析其各自的优缺点和应用范围,确定适用于本文研究的高血压临床指南应用的语言及模型。2.高血压逻辑本体库构建研究在本文中将高血压相关的病理知识提取出来加以整合,将本体的思想与临床指南知识相结合,对临床指南知识进行结构化和非结构化的表示,提取出的相关高血压概念建立领域描述逻辑本体,形成由概念、属性、实例以及相关概念之间关系组成的本体库。3.高血压医学知识表达模型设计研究结合临床医生工作流程,将高血压临床指南分解成流程图的方式。如高血压筛查、检查评估、危险分层、药物与非药物治疗、随访管理等。根据流程图化的高血压临床指南建立医学知识表达模型。4.高血压规则库整合研究利用推理引擎,将医学知识表达模型转换成符合计算机可推理要求的一系列规则,构建规则库,完成高血压医学知识库的构建。如高血压诊断规则、用药规则等。研究方法1.通过文献与网站调研,梳理本体、临床指南知识表达模型等相关理论和国内外经验做法,总结出适合高血压临床指南知识库的构建流程。2.本文采用本体开发工具Protege 4.3构建高血压临床指南本体,并参考SAGE模型进行临床指南知识表达,并将指南中的规则以流程图的方式展示,在Protege 4.3中实现本体构建、知识表达以及本体推理过程。研究结果1.主要借鉴七步法的思路,改进国际通过疾病本体模型,并参考MeSH、CMeSH主题词表的结构,对高血压临床指南进行本体构建。2.在对比传统的临床指南知识表达模型后,选择SAGE模型构建高血压临床指南知识表达模型,提高了临床指南知识表达的灵活性和科学性。3.将高血压临床指南中的诊断方法使用语义网规则语言SWRL构建成高血压SWRL诊断规则库,为高血压临床指南本体的推理提供了依据。4.使用Jess作为推理机,结合高血压临床指南本体和SWRL诊断规则库实现了高血压知识库的诊断推理,并且用一组测试实例验证了诊断推理的准确性和科学性。总结本文针对临床指南在临床应用中出现的一些障碍,提出了一种全新的临床指南数字化方法解决临床指南发展过程中的瓶颈。高血压疾病作为威胁人类健康的一种常见病,将高血压临床指南和本体研究结合起来构建高血压临床指南知识库不仅可以提高指南在临床中的利用率,也可以被临床决策支持系统直接使用。同时,高血压临床指南知识库的构建方法也适用于其它疾病临床指南知识库的构建,为临床指南构建知识库提供了一套通用的构建方法。
[Abstract]:The purpose of this study is to model the clinical guide of static text form in the form of a suitable guide expression model to form a computer understandable form, combining ontology and knowledge base technology, transforming the clinical guide of the text form into a convenient reasoning knowledge base, which helps to improve the efficiency of the clinical guide and reduce the clinical practice. On the basis of this, it can also be used as an important module of the clinical decision support system of hypertension and to reduce the difficulty of design and construction. Research content 1. comparison and analysis of the related theories of medical knowledge base construction are summarized and analyzed through the related techniques of constructing the medical knowledge base, and the design of the knowledge base suitable for hypertension is found and the design of the hypertension medical knowledge base is found. The models and techniques constructed, such as the language and tools built by the ontology, and the ontology reasoning machine, analyze their respective advantages and disadvantages and application scope, and determine the language of the application of the clinical guidelines for hypertension and the construction of the model.2. hypertension logic ontology library. In this article, the pathological knowledge related to hypertension is extracted and added. Combining the idea of the ontology with the clinical guide knowledge, the knowledge of the clinical guide is structured and unstructured, the concept of hypertension is extracted and the domain description logic ontology is extracted, and the.3. hypertension medical knowledge expression model is formed by the relationship of concepts, attributes, instances and related concepts. Design studies combined with the working process of clinicians to decompose the clinical guidelines of hypertension into flowcharts, such as hypertension screening, examination and evaluation, risk stratification, drug and non drug treatment, follow-up management, etc. The engine, transforming the medical knowledge expression model into a series of rules that conforms to the reasoning requirements of the computer, constructs the rule base, and completes the construction of the hypertension medical knowledge base, such as the rule of hypertension diagnosis and the rule of drug use. Research method 1. through literature and website research, combing ontology, clinical guide knowledge expression model and other related theories and domestic External experience, summarize the construction process suitable for the knowledge base of clinical guidelines for hypertension.2. this article uses the ontology development tool Protege 4.3 to construct the hypertension clinical guide ontology, and refer to the SAGE model for the knowledge expression of clinical guidelines, and the rules in the guide are displayed in the flow chart, and the ontology construction is realized in the Protege 4.3. 1. mainly use the ideas of the seven step method to improve the international through disease ontology model, and refer to the structure of MeSH, CMeSH theme word list, to construct the ontology of hypertension clinical guide to construct.2. after comparing the traditional clinical guide knowledge expression model, select the SAGE model to construct the clinical guide of hypertension. The knowledge expression model improves the flexibility and scientificalness of the knowledge expression in the clinical guide.3., constructs the diagnostic method in the clinical guide of hypertension by using the Semantic Web Rule Language SWRL to construct the hypertension SWRL diagnosis rule base, providing the reasoning for the reasoning of the hypertension clinical guide ontology by using Jess as a reasoning machine and combining the clinical guide of hypertension with the hypertension clinical guide. The ontology and the SWRL diagnosis rule base realized the diagnostic reasoning of the hypertension knowledge base, and verified the accuracy and scientificity of the diagnostic reasoning with a set of test cases. This paper summarized the obstacles in clinical application of clinical guide in this paper, and put forward a new clinical guide digital method to solve the development of clinical guide. Hypertension is a common disease that threatens human health. Combining the clinical guidelines of hypertension with the ontology study to construct the knowledge base of the clinical guidelines for hypertension can not only improve the utilization of the guide in clinical practice, but also be used directly by the clinical decision support system. At the same time, the construction of the knowledge base of hypertension clinical guide is constructed. The method is also applicable to the construction of knowledge base for clinical guidelines of other diseases, and provides a general method for constructing clinical knowledge base.

【学位授予单位】：北京协和医学院
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：G250.74;R544.1

【相似文献】