脂肪肝中医证候分类探究中多元统计方法应用摘要:目的:为如何选用多元统计方法用于中医证候分类研究提供参考。方法:通过2次因子分析提取脂肪肝的基本证型,利用因子得分揭示基本证型的组合规律,结合聚类分析归纳出脂肪肝常见的复合证型。结果:脂肪肝的基本证型为肝肾不足、阴虚火旺、肝郁脾虚、脾虚、脾虚湿阻、湿热内蕴、血瘀、痰湿内阻;基本证型组合形式以1〜4个多见,病机较简单时,脾虚湿阻型最常见,病机较复杂时,肝肾不足型最常见;有13.6%的患者表现为无症状型,5.8%的患者表现为不可分类型,其余患者被聚为五型,分别为脾虚痰瘀互结11.5%、肝肾阴虚18.5%、脾虚痰湿中阻32.0%、脾虚湿热内蕴10.2%和肝郁脾虚湿阻8.3%。结论:将因子分析和聚类分析等多种多元统计方法合理地结合应用,能对中医证候进行比较客观地分类。关键词:多元统计;证候分类;脂肪肝中图分类号:R259.755文献标识码:A文章编号:1673-7717(2010)01-0065-05UsingtheMethodsofMultivariateStatisticalAnalysistoStudythesyndromeCategorizationofTCMWEIHua❷Feng,LIUTao,WANGMiao,XING---本文来源于网络,仅供参考,勿照抄,如有侵权请联系删除---Lian❷Jun,ZHENGPei❷Yong,JIGuang(InstituteofDigestiveDiseases,LonghuaHospital,ShanghaiUniversityofTraditiorrnlChineseMedicine,Shanghai200032,China)Abstract:Objective:TodoreferencesforreasonablyusethemethodsofmultivariatestatisticalanalysistostudythesyndromecategorizationofTCM.Methods:Toextractthebasicsyndromepatternsoffattyliverwiththemethodoftwicefactoranalysis・Todisplaythecombinationruleofthebasicsyndromepatternswiththemethodoffactorsscore・Toconcludethemainsyndromepatternscombinedwiththemethodofclusteranalysis・Results:Insufficiencyofliverandkidney,asthenicyincausingexcessivepyrexia,liver-qistagnationandspleenasthenia,splenicasthenia,splenicastheniaanddampnessstagnation,interioraccumulationofdampness-heat,bloodstasis,interiorstagnationofphlegm-dampnessarethebasicsyndromepatternsoffattyliver・Onetofourbasicsyndromepatternscombinediscommonforfattyliver,thesyndromepatternofsplenicastheniaanddampnessstagnationis---本文来源于网络,仅供参考,勿照抄,如有侵权请联系删除---mostfrequentwhenitspathogenesisissimple,andthesyndromepatternofinsufficiencyofliverandkidneyismostfrequentwheniscomplicated・13.6%patientsofthepatternofmildsymptomand5.8%patientsbelongtothepatternofnon-categorization,otherpatientsaredividedintofivegroups:thepatternofsplenicastheniaandcoagulatetionofphlegmandstagnatedblood11.5%,yinastheniaofliverandkidneyl8.5%,splenicastheniaandretentionofphlegmaticdampness32.0%,splenicastheniaandinterioraccumulationofdampness-heat10.2%,liver-qistagnationandspleenastheniawithdampnessstasis8.3%.Conclusion:Reasonablycombinedthemethodsofseveralmultivariatestatisticalanalyses,whicharefactoranalysis,clusteranalysisandsoon,objectivecategoriesofTCMsyndromecouldbecarryout・Keywords:multivariatestatisticalanalysis;TCMsyndromecategorization;fattyliver多元统计方法是数理统计学中迅速发展起来的一个分支,是定量分析事物间复杂关系的一种综合的统计方法。它的基本思想是将高维数据投影到低维空间,找出能反映原来高维数据结构或特征的投影,从而达到研究、分析高维数据的目的。近年来多元---本文来源于网络,仅供参考,勿照抄,如有侵权请联系删除---统计方法已被广泛用于中医证候研究,在中医证候诊断与鉴别诊断方面日益体现出它的重要的应用价值。中医证候分类研究是一项复杂而艰难的工作。由于中医证候属于一个高维高阶的复杂巨系统,其内部存在大量的多重共线性、非线性及协同关系,并且,这些关系随着时间的推移发生不断的变化,所以,尽管多元统...