计算机模式识别法在类风湿性关节炎中医证型微量元素谱上的应用

计算机多因素分类法能解决单因素或双因素无法全面解决的问题。应用计算机模式识别法(多因素分类判别)对中医证候分类是个新课题。本文介绍利用此法对类风湿性关节炎的寒热错杂型、寒湿阻络型及肝肾两虚型患者分别和健康人发样微量元素含量进行t检验,变量经过处理,选取各中医证型的特征参量并进行判别,其结果为类风湿性关节炎的寒热错杂型、寒湿阻络型及肝肾两虚型和健康人的准确判别率分别为95.6％、100％和90.2％。提示可能不同中医证候的类风湿性关节炎具有不同的微量元素谱.

The Application of Computer Pattern Recognition for the Trace Element Spectrum of Different Types of Rheumatoid Arthritis

An attempt to bring forward a mahalanobis distance decision method to determine the concentration of the trace elements in healthy controls and the patients with different types of rheumatoid arthritis (RA) is prcsented. The patients are diagnosed according to TCM syndrome differentiaton, i.e., coldsyndrome associated with heat-syndr(?)me, obstruction of meridians due to coldwetness evil, and deficiency of liver-yin and kidney-yin. The result showed that the correct-diagnosis percentage of RA patients with cold-syndrome associated with heat-syndrome was 95.6%, those with obstruction of meridians due to cold-wetness evil 100%, and those with deficiency of liver-yin and kidney-yin 92%. It is suggested that the concentration of trace elements in different types. of RA patients is different.