摘要
多元分析方法是现代医学计量诊断的主要方法。本文运用多元分析方法结合电子计算机技术,对冠心病脉图计量诊断问题进行了探讨,结果:用逐步判别分析法判别冠心病及阳气虚与气阴虚不同证型,回代符合率分别为94.25%和82%.用主成分分析法的结果表明,只取三个主成分便可反映原指标中75%以上的主要信息,三个主成分基本上综合了冠心病血流动力流变学方面的病理性改变。并依据主成分值的大小对冠心病患者进行了病理分型,为临床进一步诊断与治疗提供了综合性的客观依据。
Multivariate analysis is known as the chief method of the measurement diagnosis in western medicine. This paper studied the measurement diagnosis of pulse picture for CHD by means of multivariate analysis. The results indicate that when the stepwise discriminant analysis was used to test the two types of CHD: deficiency of Yang-qi, and dificiency of qi and yin, the accuracy rates were 94.25% and 82% respectively; when the principal component analysis was used, only three principal components may reveal 75% of the chief messages from original indexes, indicating the pathologic change of CHD in hemodynamic rheology. The CHD cases were classified pathologically according to the principal component's value, which supplies a comprehensive and objective basis for the diagnosis and treatment of CHD.
关键词
冠心病
脉冲流
统计学
CORONARY DiSEASE/diagnosisPULSATILE FLOW/diagnostic useFACTOR ANALYSIS,STATISTICAL/diagnostic use
