基于决策树的检验信息系统辅助诊断功能的研究

The Study on Assistant Diagnosis of Laboratory Information System Based on Decision Tree

目的:通过构建四种常见贫血病的分类诊断模型并应用于检验信息系统(LIS),探讨如何在LIS传统功能的基础上,扩展其辅助诊断功能。方法:(1)提取在四种贫血病的分类诊断中具有诊断意义的检验指标,构建信息数据库;(2)运用决策树ID3算法构建诊断模型,将该模型通过代码在LIS中实现;(3)从LIS中提取已经确诊的四种贫血病各若干例,输入该模型验证其准确率。结果:通过临床验证,该诊断模型准确率较高。结论:可以将该模型应用到LIS扩展其辅助诊断功能。

Objective： To explore how to expand the diagnostic function of Laboratory Information System （LIS） On the basis of its traditional functions by constructing the diagnostic model of four anemias and applying it to LIS. Methods： （1）The,test indicators with diagnostic significance of the four anemias were taken out to build the information database; （2） The diagnostic model was constructed using ID3 algorithm on decision tree method and achieved in LIS; （3） Some conftrmed samples were taken out of LIS to verify the model, Results： The clinical results showed that this model has a high diagnostic accuracy rate. Conclusions： The model could be applied to extend the assistant diagnostic function of LIS.