中文摘要：

目的：从统计分析的角度探索糖尿病医案中的辨证规律。方法：应用中医四诊量表和血瘀量表,分别对138例糖尿病医案病例进行评分,并对数据进行相关统计分析。结果：①聚类分析将筛选后的四诊因子分为了肾阳虚类和肾阴虚类。②四诊与血瘀19因子比较发现舌诊因子在糖尿病诊断中权重较高。结论：应用聚类分析的统计方法在一定程度上客观地验证了糖尿病中医辨证规律。

英文摘要：

Objective：Explore the syndrome differentiation principle of diabetic cases from the perspective of statistical analysis.Methods：Employ the Four-Diagnoses Scale and Blood Stasis Scale of traditional Chinese medicine（TCM） to give the 138 medical cases marks and analyze the data.Results：1.The four-diagnoses factors are classified as kidney yang deficiency class and kidney yin deficiency class by cluster analysis.2.Tongue diagnosis is the most important factor among the 19 factors both in four-diagnoses and blood stasis.Conclusion：To some extent,the TCM differentiation principle of diabetes can be proved by cluster analysis.