基于隐树模型的40岁～65岁妇女原发性骨质疏松症中医证候要素分析  

Latent tree model analysis on TCM syndrome factors of primary osteoporosis among 40～65 years old women

作　　者：谢雁鸣[1] 杨伟[2] 田峰[1] 易丹辉[2] 虞鲲 康树[4] 刘峘[1] 李建鹏[1] 支英杰[1] 

机构地区：[1]中国中医科学院中医临床基础医学研究所,北京100700 [2]中国人民大学统计学院,北京100872 [3]上海市大华医院,上海200237 [4]北京中医药大学东直门医院,北京100700

出　　处：《中国中医基础医学杂志》2011年第7期731-734,共4页JOURNAL OF BASIC CHINESE MEDICINE

基　　金：国家自然科学基金面上资助项目-"基于非齐性markovmodel建立病证结合的绝经后骨质疏松证早期风险评估模型"(30873339);中国中医科学院第二批自选选题项目(Z02102)

摘　　要：目的:运用机器学习方法探索40岁～65岁妇女原发性骨质疏松症的中医基本证候要素特征。方法:运用隐类分析方法对北京和上海社区骨质疏松人群中医证候相关数据进行分析并建立隐树模型。结果:得到BIC评分为-22950.125的隐树模型,模型中的隐变量展现了骨质疏松人群以肾虚、肝虚、阳虚、阴虚、血瘀等为主的基本证候要素特征。结论:隐树模型可以客观化、定量化地揭示中医症状间的复杂关系,为中医证候的定量化研究提供研究思路。Objective：To explore the TCM syndrome factors of women aged 40~65 years with primary osteoporosis（OP） through machine learning methods.Methods：Latent class analysis was applied to establish a latent tree model based on TCM syndromes related data of OP population in Shanghai and Beijing communities.Results：A latent tree model with the highest BIC score-22950.125 was constructed and the latent variables showed in the model indicated the basic TCM syndrome essential factors of POP,including kidney deficiency,liver deficiency,yang deficiency,yin deficiency,kidney essence insufficiency and blood stasis.Conclusion：Latent tree model can be used to reveal the complex relationship between TCM symptoms,which provids a new method for quantitative and objective research of TCM syndromes.

关 键 词：隐树模型 原发性骨质疏松症 中医证候要素 

分 类 号：R274.9[医药卫生—中西医结合]