基于数据挖掘技术的焦虑抑郁共病中医证候学规律研究

项目名称： 基于数据挖掘技术的焦虑抑郁共病中医证候学规律研究

项目编号： No.81202621

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 医学八处

项目作者： 孙文军

作者单位： 北京中医药大学

项目金额： 23万元

中文摘要： 焦虑抑郁共病是指抑郁与焦虑均符合诊断标准的一种共病状态，具有发病率高、易复发、病情重的特点。中医药对于本病的治疗有独特的优势。现有的研究对于证候分型及治疗方法，多从主观的个人经验出发，结论不够严谨，缺乏科学性和代表性。目前本病的证候规律、治疗原则、方法等方面尚缺乏统一认识，极大地限制了中医药治疗GAD的研究。 贝叶斯网络技术、聚类分析、关联规则和Logistic回归是先进的数据挖掘技术。它们能够客观地、规范地建立症状之间、证候要素和证候靶位之间、证候类型与症状之间的逻辑关系，提取证候规律，为最终确立证型提供依据。 我们首先采用规范的DME方法进行大样本的中医证候学相关指标的临床观察，然后运用贝叶斯网络、聚类分析、关联规则、Logistic回归分析等多种数据挖掘技术建立焦虑抑郁共病的证候模型，总结证候学规律，结合专家经验和中医理论，制订《焦虑抑郁共病的中医证候诊断标准和治疗方案（草案）》。

中文关键词： 焦虑抑郁共病；数据挖掘；证候；诊疗方案；贝叶斯网络

英文摘要： Comorbid anxiety and depression disorder(CAD) is combidity of both anxiety and depression. CAD has a high disease incidence. It's easy to relapse, and difficult to be cured. TCM has some advantages in this field. However, past research mostly define the syndrome patterns and therapy formula from their personal experiences, and the conclusions are not strict and reasonable enough. Researchers now still have not form a criterion in syndrome patterns, principle of treatment and therapeutic methods of CAD. The current situation obstructs the research of CAD in TCM. Bayes Net, clusteranalysis, association analysis and logistic regression is advanced methods in data mining field. It can build an objective and critical logistic relationship between symptoms, syndrome elements, symptoms and syndrome, find syndrome discipline without predetermined syndrome patterns, and provide proof for determining syndrome patterns. The research will first observe the symptoms of large sample GAD patients in critical DME method, and analyze the data with Bayes Net, cluster analysis association analysis and logistic regression, finally discover the syndrome discipline of CAD, formulate the diagnostic criterion and therapeutic formula of CAD by TCM.

英文关键词： Comorbid Anxiety and Depression；Datamining；Syndrome；Treatment Protcol；Bayes Net