论文部分内容阅读
目的:探讨Bayes判别分析方法在慢性再生障碍性贫血(CAA)中医辨证分型中的应用,为其规范化中医证候诊断提供科学依据.方法:利用“全国中医医疗与临床科研信息共享系统”收集我院372例CAA患者症状评估量表信息,涵盖CAA常见的12种症状.将患者中医辨证分型分为肾阳虚型、肾阴虚型以及肾阴阳两虚型3组,以Bayes判别分析方法建立CAA辨证分型的判别方程,并采用自身验证回代法对函数式进行检验.结果:根据12项症状建立CAA中医证型判别函数,回代法判别肾阳虚证的正确率为94.93％,肾阴虚证的正确率为59.26％,肾阴阳两虚证的正确率为82.81％,总体正确率为88.17％.结论:将判别分析等多元统计方法用于中医证候的研究中,有利于提高临床诊断的标准化、客观性以及准确性.“,”Objective:To explore the application of Bayes discriminant analysis method used in traditional Chinese medicine (TCM) syndrome diagnosis of chronic aplastic anemia (CAA),in order to provide the scientific basis for the standardization of TCM syndrome diagnosis.Method:We used the “TCM and clinical research combined information system” to collect the data of 372 CAA patients.All the 372 patients were divided to kidney-yang deficiency group,kidney-yin deficiency group and kidney-yin-yang-deficiency group,then the “CAA Symptom Assessment Scale” information which covers 12 kinds of common symptoms were selected.The discrimination equations of CAA TCM types were set up by Bayes discriminant analysis method,and we used back substitution method to verify their own functional test.Result:CAA TCM type discriminant functions were established according to the 12 symptoms.The accuracy of kidney-yang deficiency was 94.93％,the accuracy of kidney-yin deficiency was 59.26％,the accuracy of kidney-yin-yang deficiency was 82.81％,and the overall accuracy was 88.17％.Conclusion:Discriminant analysis would be very good application in the research on TCM syndrome diagnosis.