摘要
目的为建立多发性硬化的中医证候量化诊断标准做初步研究。方法从四诊信息出发,收集多发性硬化患者500例,以基于熵的复杂系统划分方法提取多发性硬化中医症状条目组合以及归纳中医证候要素组合,以多元回归方法确立各个症状条目对各相应中医证候要素组合的贡献度,以诊断性试验ROC曲线建立诊断阈值。结果熵聚类结果得到多发性硬化疾病本身特点的6组证候要素集合,分别为肝肾阴虚、脾肾阳虚、脾气虚、血瘀、痰湿热、动风。不同中医症状条目对相应证候要素组合诊断贡献度不同,各证候要素组合有不同的诊断阈值,经过与专家辨证为金标准的回顾性检验,具有良好的灵敏度和特异度。结论各统计学方法综合运用建立起的《多发性硬化证候要素诊断量表》具有良好的诊断效能,符合临床使用要求。
Objective To study on establishing quantized criteria for TCM syndromes of multiple sclerosis. Methods The patients （n = 500 ） were chosen according to TCM four examinations. The combinations of TCM symptom entries and combinations of TCM syndrome factors were extracted by using entropy-based complex system partition method, and contribution degrees of all symptom entries to corresponding combinations of TCM syndrome factors were determined by using multiple regression method. The diagnostic thresholds were established with ROC curves of diagnostic test. Results There were 6 combinations of TCM syndrome factors got from entropy clustering results, and they were liver- kidney yin deficiency, spleen-kidney yang deficiency, spleen qi deficiency, blood stasis, phlegm damp- ness and phlegm heat and engendering wind. Different symptom entries had different contribution degrees to corresponding combinations of TCM syndrome factors, and all combinations of TCM syndrome factors had different diagnostic thresholds, which had higher sensitivity and specificity through the retrospective verification taken experts＇ syndrome differentiation as the gold standard. Conclusion The Syndrome Diagnostic Scale of Multiple Sclerosis established by applying various statistical methods has good diagnostic performance and conforms to clinical demands.
出处
《北京中医药大学学报》
CAS
CSCD
北大核心
2014年第1期68-72,共5页
Journal of Beijing University of Traditional Chinese Medicine
关键词
多发性硬化
证候要素
数据挖掘
量表
multiple sclerosis
syndrome factors
data mining
scale
