基于数据挖掘与运气学说的中医肺系疫病预测与辨治规律研究

With the full Collection of pulmonary system epidemic literatures from ancient medical books and medical journals, we try to extract diseases that have clinical significence in Jiangsu and Zhejiang(influenza A/measles/pulmonary tuberculosis/SARS). The results will be sorted by cause of disease, name of disease, symptom and principle of treatment respectively. After necessary cleaning, integration as well as terminology standardization, these data will be processed and be constructed as an epidemic data warehouse.The factors of yun-qi theory will be quantitative expressed at meanwhile,and we will analyze the correlation with meteorological data and disease transmission,in addition,a prediction model of infectious diseases will be built. On which it based, techniques such as multivariate analysis and Scale free network analysis will be introduced to find out valuable informations about complex relationships of various pathogenic factors of epidemic, such as season, environment, weather, gender, age, ancient and modern name of disease, cause and symptom, treatment as well as the property, function, dosage, pharmacological effect of drugs, which will help to construct a clinical decision making system of epidemic. The ultimate goal of our research is to provide modern chinese medicine new train of thought and methodology to the prevention and treatment of acute epidemic diseases occured nowadays or may occur in the future.

对古代医药著作中治疗疫病及近现代中医药治疗呼吸系统急性传染病的临床报道资料进行全面收集与整理，选择江浙地区具有较大临床意义的病种（甲型流行性感冒、麻疹、肺结核、重症急性呼吸综合征），按五运六气、发病因素、病名、病状、治则、用药等进行分类，然后对数据样本进行预处理，通过数据清理、数据集成、术语规范等，建立中医疫病文献数据仓库。同时将运气学说的运气因子量化表达，分析其与气象数据及传染病流行的相关性，并建立传染病预测模型，同时利用多元分析技术，结合无尺度网络分析方法等研究疫病发病季节、环境、气候、性别、年龄、病名、病因与证候、治则、药物及其药性、功能、剂量、药理作用等之间的相关性，寻找中医疫病辨治用药规律，并将之与当代治疗急性传染病的辨治思路与用药原则进行对比研究，结合预测的结果为每年的疫病治疗提供有效预防和治疗方案。

课题以江苏省南京地区为例，基于运气学说对中运、司天、在泉与气候变化的相关认识，利用降水量、温度、风速、相对湿度等气象指标距平反映南京地区近12年气候变化特征，探讨其与运气推演的“中运”“司天”“在泉”时段的符合情况，证明运气学说在南京地区有一定的适用价值，并能以此为基础开展肺系疫病预测研究。对古代医药著作中治疗疫病及近现代中医药治疗呼吸系统急性传染病的临床报道资料进行全面收集与整理，选择具有较大临床意义的肺系疫病（麻疹、肺结核等呼吸道传染病），运用SPSS软件对近年来肺系疫病高发情况进行统计学分析，发现疫病发病情况与运气各时段存在一定的差异性，有助于指导疫病的预防。采用齐夫定律、Apriori算法、改进互信息法、复杂系统熵聚类、无监督的熵层次聚类等多种数据挖掘手段，区别古今疫病治疗辨治差异，发现现代肺系疫病治疗中清热药的使用十分突出，而古代则多采用解表宣散法。此外，疫病的外用治疗原则应以祛邪为主，扶正补虚的运用相对较少。