基于复杂网络方法的现代名老中医诊治COPD 证-方-药规律研究

Traditional Chinese medicine (TCM) has beneficial effects on chronic obstructive pulmonary disease patients (COPD), especially prominent veteran TCM physicians’ rich experiences for COPD, however, the experiences lack of systematically arrangement and data mining. Based on the established database of modern prominent veteran TCM physicians’ experience in lung disease and database of TCM for lung diseases indexed from academic journals, the following study will be carried out: to begin with, collating the database of the prominent veteran TCM physicians’ experience in syndromes-prescriptions-drugs for COPD, then multivariate statistics analysis methods complex network methods and association rules analysis will be employed to evaluate syndrome elements and distribution of syndromes, and to find the regulations of COPD etiology and pathogenesis. In addition, the networking model of syndromes-prescriptions-drugs for COPD will be structured on the basis of COPD common syndromes, and the common law and individual differences and the topological characteristics will be analyzed from the complex network indexes: distribution of complex networks, correlation degree, and the network interface number, average distance, cluster coefficient. Finally, the multi-level classification of syndromes-prescriptions-drugs for COPD will also be built by overlapping community discovery algorithm based on node, label propagation algorithm based on link and local optimization algorithm based on module. The established complex network methods will provide the inspirations for other disease researches, and lay the foundation for the conversion application of modern prominent veteran TCM physicians’ experience. Through the study, the key technology and suitable method will be optimized to provide methodology reference for the systematically arrangement and data mining, and provide technical support for the deep research of the basic theories of lung disease.

中医药治疗COPD效果显著，名老中医经验尤为丰富而鲜活，但缺乏系统整理与挖掘。基于已建立的《现代名老中医肺病数据库》和《期刊中医肺病数据库》，拟研究内容为：①建立《现代名老中医诊治COPD的文献研究数据库》；②运用多元统计方法、神经网络及关联规则分析COPD证素、证候分布特征,揭示COPD病因病机、证候规律；②基于COPD常见证候，构建COPD现代名老中医证-方-药信息的网络化模型，从复杂网络的节点度及度分布、度的相关性、网络介数、平均距离、集聚系数等对COPD证-方-药配伍的共性规律与个体差异进行研究，分析其复杂网络的拓扑结构特征；③采用基于节点的重叠社团发现算法、基于链接的标签传播算法及基于模块度的局部优化算法进行COPD证-方-药配伍的多层次的类别划分。为同领域其它研究提供示范，为名老中医经验转化奠定基础。通过研究以优化适宜的关键核心技术方法，为研究现代名老中医经验提供方法学参考。

中医药治疗慢性阻塞性肺疾病（COPD）效果显著。名老中医经验尤为丰富而鲜活，但缺乏系统整理与挖掘。基于已建立的《现代名老中医肺病数据库》和《期刊中医肺病数据库》，研究内容为：①建立《现代名老中医诊治COPD的文献研究数据库》；②运用频次、频率分析COPD证素及组合规律、证候分布特征，揭示COPD病因病机规律；③运用关联规则初步分析COPD证-方-药规律；④基于COPD常见证候，构建COPD现代名老中医证-方-药信息的网络化模型，从复杂网络的节点度、中心性、平均距离、集聚系数等方面，对COPD证-方-药配伍规律进行研究，分析其复杂网络的拓扑结构特征；⑤采用基于派系过滤CPM方法的重叠社团算法及基于局部优化算法进行COPD证候-药物及证素-药物配伍的多层次的类别划分。结果显示：采用频率及关联规则，得到COPD常见证候6种，即：痰湿阻肺证、痰热壅肺证、肺脾气虚证、肺肾气虚证、肺肾气阴两虚证、血瘀证。以痰湿阻肺证为例，采用频数及中心性分布，得到该证核心方剂及核心中药，方剂有二陈汤、三子养亲汤、小青龙汤、苏子降气汤、苓桂术甘汤。方剂常以合方的形式出现，如：二陈汤合苏子降气汤或苓桂术甘汤或三子养亲汤。核心中药有半夏、茯苓、陈皮、杏仁、苏子、甘草、五味子、白术、桂枝、麻黄。采用边权重和关联规则，得到中药配伍，如半夏、茯苓、陈皮；麻黄、杏仁、紫苏子；紫苏子、前胡等。采用派系过滤CPM方法，得到中药的多层次类别划分，并运用Gephi0.9.2软件形成可视化图谱，构建COPD证候-中药配伍，如对肺肾气虚证中复杂网络节点度≥40的53味中药进行划分，得出5个社团。该研究为同领域其它研究提供示范，为名老中医经验转化奠定基础。通过研究以优化适宜的关键核心技术方法，为研究现代名老中医经验提供方法学参考。