基于隐结构结合灰色关联的名老中医辨治哮喘临床诊疗策略研究

Chinese medicine treatment of bronchial asthma effect is remarkable, the old Chinese medicine clinical diagnosis and treatment strategy is the bright pearl of the treasure house of traditional Chinese medicine (TCM), but the lack of system arrangement and mining. Database based on the modern name old doctor of traditional Chinese medicine lung disease and pulmonary disease databases of journal literatures of traditional Chinese medicine, proposed the research content as follows: ①In order to establish the modern structured database name old doctor of traditional Chinese medicine treatment of bronchial asthma. ② Using the method of implicit structure similarity score - grey correlation analysis, in turn, the method of progressive mining of old Chinese medicine diagnosis and treatment of bronchial asthma clinical diagnosis and treatment strategy; ③Data hidden structure model was constructed based on symptoms, with mutual information and information coverage of quantitative hidden variables and the correlation between variables, using implicit class probability and conditional probability depict the distribution of hidden variables and show, extract the causes of asthma, syndrome elements, evolution and pathogenesis, syndrome classification; Constrained data sets with similarity algorithm, and carries on the classification, reuse implicit structure perfect syndrome classification method, analysis based on syndrome differentiation; Finally with grey relation and association rule mining, square, relationship between drug, summarizes the legislative formula and sent the clinical features. ④This provides demonstration for other research with field, experience in the name of the old Chinese medicine syndrome differentiation of manifest lay the foundation. Through the study of key core technology to optimize the appropriate method, for the modern name old doctor of traditional Chinese medicine clinical diagnosis and treatment strategy of mining provide methodology reference.

中医药治疗支气管哮喘效果显著，名老中医的临床诊疗策略鲜活而丰富，但缺乏系统整理与挖掘。基于《现代名老中医肺病数据库》和《中医肺病期刊文献数据库》，拟研究内容为：①构建《现代名老中医辨治支气管哮喘结构化数据库》；②运用隐结构法-相似性评分-灰色关联依次递进挖掘的方法揭示名老中医辨治哮喘的临床诊疗策略；③基于症状数据构建隐结构模型，用互信息和信息覆盖率量化隐变量间及其与显变量间关联度，用隐类概率和条件概率刻画隐变量和显变量分布情况，提取哮喘的病因、证候要素、病机演变、证候分类；再结合证型用相似度算法约束数据集进行数据分类，复建隐结构模型，完善证候分类，分析辨证依据；最后结合灰色关联和关联规则挖掘证、方、药物间关联关系，总结立法组方思路和遣方用药特色。④为同领域其它研究提供示范，为名老中医辨治经验的显化奠定基础。通过研究以优化适宜的关键核心技术方法，为现代名老中医临床诊疗策略的挖掘提供方法参考。

支气管哮喘是肺系多发性疾病，名老中医临床诊疗策略是中医药学宝库中的璀璨明珠。本项目通过系统整理和挖掘名老中医辨治哮喘的医案，先构建《古今名老中医辨治支气管哮喘结构化数据库》，以隐结构法研究哮喘的病因、病机与证候分布特征为基础，采用相似性评分、灰色关联、复杂关联网络等探究诊治哮喘的立法组方思路和遣方用药特色。结果表明，4097例哮喘古籍医案前60位症状和60味中药建模得到39个隐变量，100个隐类，10个综合聚类模型，其中Z10的阈值7.5最高，Z8咽痛信息覆盖度71%最高，Z2的前胡分值18.5最高。二分类Logistic回归模型拟合出不同病性和5类证型的相关关系，其中气虚和脾肺气虚证的相关性Exp (B)值最高达73494.287。从证候的频数分布、以症推证、以药（方）测证得出喘病的常见证候有14种，新发现的证候有肺气阴两虚、脾肾阳虚、脾胃虚弱证。哮喘常用的中药偏辛、甘、苦味，归肺、脾、胃、肾经，常用具有补气、止咳平喘、发散风寒、理气、清化热痰、利水消肿功能的中药；常用药组有甘草、半夏、紫苏和陈皮，半夏、茯苓和陈皮，陈皮、大腹皮和紫苏。常用的症-药关联有气喘、咳嗽分别与人参和地骨皮、陈皮和茯苓，咳嗽、咯血和杏仁、人参。5397例近现代哮喘医案挖掘表明针对病性风、寒宜用麻黄、桂枝、细辛、甘草、射干、连翘；病性为气虚、阴虚宜用麦冬、沙参、熟地黄、五味子、太子参；病性痰、热宜用鱼腥草、石膏、黄芩、葶苈子、桑白皮；病性风、热，宜用连翘、薄荷、荆芥、桑叶、竹叶、蝉蜕；病性痰、湿，宜用桂枝、细辛、麻黄、杏仁、白术、茯苓、厚朴、陈皮、白芥子、紫苏子；病性寒、湿，宜用藿香、厚朴、枳实、薏苡仁、苍术、羌活、黄芪。本研究为哮喘辨治规范的修订和完善提供理论依据，同时为实现名老中医诊疗经验的传承和利用提供方法学参考。