Implementation Of Latent Variable Models Based On Hypertension Syndrome Elements In R

Objective:The main purpose of this study is to investigate the relationship between the four diagnostic information of TCM about hypertension with hypertension syndromes and syndrome elemets.These models about syndromes and syndrome elements were made by R language.First,to explore the TCM syndromes' classification,and to confirm the structure of TCM syndromes.Then to evaluate the rationality of four diagnostic information's grade in each syndrome,and to extract elements of TCM Syndromes.The establied methods can provide reference for the quantitative study of syndrome differentiation of traditional Chinese medicine.Method:Application of R software for hypertension in TCM diagnosis information,firstly,using exploratory factor analysis and confirmatory factor analysis on hypertension syndrome rough classification,and then the second-order confirmatory factor analysis of rough classification results in Yin deficiency with dampness syndrome and verified by adjusting the load,then the item response theory evaluation the rationality of clip dampness syndrome four diagnostic information level of yin deficiency.Finally,the potential categories of yin deficiency and dampness syndrome elements were analyzed by latent class model.Result:Factor analysis identified the main TCM Syndromes of hypertension can be divided into five categories,respectively:hyperactivity,hyperactivity of liver fire,yin deficiency of liver and kidney,phlegm,yin deficiency and dampness.The second-order confirmatory factor analysis showed that yin deficiency with dampness syndrome contains 3 elements of syndrome,syndrome factor 1 contains 8 indicators,6 indicators(dry eyes,eye itching,forgetfulness,tinnitus,blurred vision,Yingfengliulei)standardized coefficient is higher than 0.4,2 syndrome elements including 3 indexes 2,the index(urine after more than Lek,poor urine)standardized coefficient is higher than 0.4,3 syndrome elements include 9 indicators,6 indicators(body weight,urine clear long,sticky mouth,edema,nocturnal polyuria,numbness of limbs)standardized coefficient is higher than 0.4.In the syndrome of yin deficiency and dampness,the representativeness of the other 18 indexes is better than that of the tongue and the pulse.Foot pain only need to divide three grades(no,light,heavy)can be divided into four limbs numbness of the 2 grades(with,without)level can be,the rest of the 16 indicators are more reasonable classification.The probabilities of the 4 categories corresponding to the latent variables were 0.3202,0.1044,0.1022 and 0.4732,respectively.The conditional probability of visual ambiguity and forgetfulness in the 4 categories was higher and appeared in the top 5.Conclusion:The 3 elements of syndrome of Yin dampness syndrome:syndrome factor 1 can be interpreted as a disease in the liver,disease of the deficiency syndrome factor 2 can be explained in kidney disease,disease of yin deficiency syndrome factors 3 can be interpreted as a disease of wet.There is a high degree of correlation between the syndrome of yin deficiency and dampness and the 20 indexes selected.Grade response theory can better evaluate the rationality of the four diagnostic information in TCM syndrome classification.The potential categories for TCM syndrome classification of multiple two categorical response variables have good results.