数据挖掘技术在建立肺结核病单病种费用模型中的研究

肺结核病具有发病率高、危害严重、消耗医疗资源大的特点。因病情程度和并发症等差异，其住院费用的差别也较大。本文利用数据挖掘技术，针对南京市胸科医院信息系统中所积累的医疗病历数据及病人医疗费用数据，构建了相应的病人医疗费用数据库。采用决策树、神经网络等数据挖掘方法，对肺结核单病种医疗费用数据进行分析、处理，建立了用3层10分类的决策树表示的费用模型，比较客观地反映了历史数据中所蕴涵的相关知识或规律。

Research on data mining technology in establishing pulmonary tuberculosis single disease cost model

Pulmonary tuberculosis has features of high incidence, serious harm, consumption of medical resources. Due to the differences in severity and complications, the difference of its cost of hospitalization is also large. In this paper, using the data mining technology, the medical record data and patient medical expense data accumulated in the information system construction of NanJing Chest Hospital, the patient medical expenses to the corresponding data warehouse were established. Using decision tree, neural network, data mining methods, analysis, treatment for PTB medical expense of single disease data were analyzed and cost model represented by decision tree classification of 3 layers 10 was established. The related knowledge or rules contained in historical data were objectively reflected.