融合决策树模型的体检套餐定制系统在军队大型医院的应用  

Medical Package Custom Method with Decision Tree in Large-scaled Military Hospital

作　　者：刁丽萍[1] 王翊[2] 吴军[1] 孟祥贵[1] 

机构地区：[1]第三军医大学新桥医院健康管理科,重庆400037 [2]重庆大学计算机学院

出　　处：《华南国防医学杂志》2014年第12期1212-1214,共3页Military Medical Journal of South China

基　　金：重庆市自然科学基金(CSTC2010CF5024)

摘　　要：目的利用数据挖掘的思想发掘体检套餐定制过程中有用的知识,构建智能体检套餐定制系统。方法根据国际疾病分类标准ICD-10将体检者疾病分类,然后使用C4.5算法构建决策树,利用这些数据构建一种简单的体检定制系统模型。结果以2型糖尿病为例,采用500例体检者的体检信息建立决策树模型,运用该决策模型诊断糖尿病准确率高达85.71%,误诊率仅为0.64%。结论实验表明融合决策树模型的体检套餐定制系统表现出较好的性能。Objective Important knowledge in medical package custom was discovered by theory of data mining,and structruing intelligent health check medical package custom system was structured.Methods ICD-10 was used in the classification of disease of volunteers,and a decision tree was constructed by the C4.5 algorithm.Basing on these data,a simple customized physical examination system was structrued.Results In the experiment,examination information of 500 hundreds 2-diabetic are used to construct a debate decision model.Basing on this model,the diagnostic accuracy rate was up to 85.71% and the misdiagnosis rate was only 0.64%.Conclusion Experimental results suggest that the medical package custom method with decision tree performs well.

关 键 词：决策树 体检套餐定制 C4.5算法 

分 类 号：TP311.132.1[自动化与计算机技术—计算机软件与理论]