【作者】 陈楚湘；

【导师】 沈建京；

【作者基本信息】 解放军信息工程大学 ， 计算机软件与理论， 2011， 博士

【摘要】 当前,信息技术和智能方法高速发展,其在各行业、各领域的应用极大地推动了行业的标准化和规范化建设。中医作为中华医学的瑰宝,经过数千年的医疗实践,逐步形成并发展成为具有独特的医学理论体系,丰富的治疗经验的完整医学。但作为经验医学,缺乏中医诊疗标准,即证候诊断标准和疗效评价指标等客观标准。因此,以中医临床数据为基础,综合运用人工智能和系统工程的理论与方法,开展中医证候诊断标准和疗效评价方法的基础理论研究,建立中医辨证论治的证候诊断标准和疗效评价指标体系,进而实现中医证候语义的形式化表达和面向服务的分布式中医诊疗应用集成,具有重要的学术意义和应用价值。本文基于上述问题,以国家重点基础研究发展计划973计划课题——基于肺炎的辨证论治疗效评价方法基础理论研究为支撑,围绕中医诊疗标准,研究了人工智能领域相关智能方法和应用技术。主要工作是：1.提出了R_Apriori算法,为实现从大量临床数据中挖掘出中医证候诊断标准和疗效评价指标体系提供高效实用的方法支撑。2.综合运用R_Apriori算法、神经网络算法、粗糙集理论和AHP方法,给出了中医证候诊断标准和疗效评价指标体系建立的一般过程,并以临床数据为基础,建立了中老年肺炎中医证候诊断标准和疗效评价指标体系。3.建立了中医证候本体,按照概念的层次结构描述了中医证候本体的概念体系,提出了包括基本公理、诊断公理和疗效评价公理在内的中医证候本体的公理体系；4.对中医诊疗服务进行语义标注和语义推理,并建立了基于本体和Web服务的中医证候挖掘和诊疗系统。主要创新点有：1.提出了改进关联规则算法—R_Apriori,提高了数据挖掘的运行效率和结果的有效性。2.综合运用多种智能处理方法,建立了中老年肺炎的中医证候诊断标准和疗效评价指标体系。3.提出了中医证候本体的概念,建立了中医证候本体的概念体系和公理体系。4.将中医诊疗过程以细粒度的Web服务进行语义标注,采用基于描述逻辑和规则相结合的语义推理方法,实现了中医诊疗服务的语义推理。本文所采用的方法和得到的诊疗标准在相关医院的实际应用中得到了验证,为973计划课题提供了理论和方法上的支撑。将智能处理方法运用于中医诊疗标准的建立,较好地解决了中医标准化问题,开拓了中医诊疗标准制定的新思路,为中医概念的规范化提供了参考；建立面向中医从业人员的研究平台和面向大众的医疗服务平台,有利于中医知识的获取和共享,为中医的传承提供科学的依据。更多还原

【Abstract】 At present, information technology and intelligence methods have gained rapid development, and they facilitate greatly the standardized and normalized process for their broad applications in many professions and fields. As a treasure of Chinese medical science, Traditional Chinese Medicine(TCM) has a history of thousands of years, and has been gradually formed a kind of medical science with a unique architecture of medical theory and rich experience on treatment. However, for its high dependency on experience, TCM is still lack of the criteria for diagnosis and treatment, such as the criteria for syndrome diagnosis and the indicator architecture for evaluation of treatment efficacy. Aiming to solve the problem mentioned above, the author adopts AI and systems engineering as the method of research, tries to dive into foundational theory about the criteria for TCM diagnosis and the method for evaluation of treatment efficacy. It’s important for TCM to establish the criteria for diagnosis and the indicator architecture for evaluation of treatment efficacy.By this means, it can be easily done that the formalized representation of the semantics in TCM syndrome and the integration of service-oriented TCM applications on distributed platforms. Supported by an issue of 973 plans with title of "Foundational Research on Dialectical Evaluation for the Efficacy of Pneumonic Diagnosis and Treatment", this dissertation makes a deep research into the intelligence methods and application technologies on the establishment of criteria for TCM diagnosis and the indicator architecture for evaluation of treatment efficacy.The main work can be summarized as following four aspects:1. Proposes an algorithm namely R_Apriori, which provides a way for establishing the criteria for TCM syndrome diagnosis and the indicator architecture for evaluation of treatment efficacy from mass clinical data.2. Provides an ordinary process for the establishment of the criteria of TCM diagnosis and the indicator architecture for evaluation of treatment efficacy, and then, takes pneumonia in middle-aged and old patients as an example, establishes the criteria of TCM pneumonic diagnosis and the indicator architecture for evaluation of pneumonic treatment efficacy.3. Constructs the ontology for TCM syndrome and its hierarchically conceptual structure, and proposes an axiom architecture which consists of basis axioms, diagnostic axioms and evaluation axioms for treatment.4. Implements semantic annotation and semantic reasoning on the services of TCM diagnosis and treatment, and builds an application system for the mining of TCM syndrome and the treatment based on ontology and Web services.The innovation of this dissertation can be concluded as the following four points:1. Proposes an improved algorithm for associated rules namely R_Apriori, which would promote the running efficiency of data mining, and guarantee the validity of results.2. Builds the criteria for diagnosis of TCM syndrome and the indicator architecture for evaluation of TCM treatment efficacy, with the help of various intelligent methods.3. Proposes the concept of TCM syndrome ontology, and builds the hierarchical structure and the axioms for the ontology.4. Provides semantic annotation for the fine-grained Web services of TCM diagnosis and treatment, and then, semantic reasoning of these services can be realized with the help of both description logics and rules.The following results can be concluded from this research and its applications:The methods and the criteria proposed by this dissertation have been verified in practical applications from many hospitals, which provide theoretical and methodological support for 973 research subject. It would be turned out that putting intelligent methods into the establishment of criteria for TCM diagnosis, can solve the problem of normalization, and it provides a new idea for the criteria for TCM diagnosis and reference to standardization of TCM conception. Through building a service-oriented applications integration distributed platforms towards both medical employees and the public, TCM knowledge can be acquired and shared, and scientific basis can be provided for TCM transferring.更多还原