中医肝病电子病历病证关联分析探索性研究

Exploratory Research on Association Rules about Electronic Medical Records of TCM Liver Disease

目的:探讨数据挖掘技术中的关联规则与决策树分类在中医肝病诊断中的应用。方法:整理筛选来自某中医医院肝病科电子病历的数据,利用Weka数据挖掘软件对其进行数据分析,主要采用关联规则中的Apriori算法和决策树分类中的C4.5算法。结果:经算法分析产生十条相关规则,并生成中医常见肝病的决策树,揭示了检查指标与中医辨证间的关系。结论:数据挖掘结果与医生实际诊断结果基本一致,可将该方法用于基于电子病历的疾病诊断,具有一定实用价值。

Objective: To investigate the application of the association rule and decision tree classification of data mining technology on diagnosis of TCM liver diseases. Method: To collate and screen the data about hospitalization medical records of TCM liver disease, and use Weka to carry on the data mining analysis, which mainly uses Apriori algorithm in association rules and C4.5 algorithm in decision tree classification. Result: Ten rules were found and the decision tree of the common TCM liver disease was generated. The relationship between lab examination information and TCM dialectical treatment has been given. Conclusion: The results obtained from the data mining are generally consistent with the results from the doctors. This method can be used in disease diagnosis based on the electronic medical record, which has certain practical value.