一种基于数据挖掘的中医定量诊断方法

A quantitative diagnostic method based on data-mining approach in TCM

目的 应用数据挖掘技术开展定量诊断研究。方法 应用贝叶斯网络方法通过分析 474 例血瘀证临床诊断数据进行血瘀证定量诊断。结果 该方法发现了血瘀证的7个关键症状,并定量计算其诊断贡献度。基于这些关键症状建立的简单贝叶斯分类器模型对血瘀证诊断的准确率达到96.6 %。结论 贝叶斯网络方法在中医定量诊断中具有良好的应用前景。

Objective To carry out a study by using data-mining techniques in TCM quantitative diagnostics for promoting normalization and standardization of TCM diagnoses. Method Bayesian network method was used to make the quantitative diagnosis of blood stagnation syndrome based on TCM clinical diagnostic data of blood stagnation syndrome collected in 474 cases.Results 7 Key symptoms of blood stagnation syndrome were found out by using this method and their contribution degrees were calculated quantitatively. The rate of accurateness in diagnosing blood stagnation syndrome by a na?ve Bayesian classifier model based on these key symptoms was proved to be 96.6%.Conclusion The results suggest that the Bayesian network method has a promise of good future in making TCM quantitative diagnoses.