基于数据挖掘方法的冠心病心绞痛合并糖尿病患者理化指标与气阴两虚证诊断模式的建立

Diagnostic Mode Establishment of Physicochemical Indexes and Dual Deficiency of Qi and Yin Pattern Based on Data Mining Method in Patients with Coronary Heart Disease and Angina Pectoris Combining with Diabetes Mellitus

目的:拟通过初步数据处理和深度数据挖掘的方法筛选出同冠心病心绞痛合并糖尿病(CHD&DM)气阴两虚证最为相关的理化指标,建立理化指标与证候要素的诊断模式。方法:将临床收集的100例CHD&DM患者的临床基本资料、中医四诊信息及理化指标进行综合分析。在t检验、非参数分析、Spearman相关分析的基础上,进一步采用决策树的数据挖掘方法探讨气阴两虚证与理化指标之间的对应性关系。结果:通过决策树的数据挖掘方法得出由射血分数、LDLC、2hpG、QRS时限、血清P、TG共6个理化指标组成的气阴两虚证决策树模型,该模型共形成7条气阴两虚证识别途径,检测准确率高达77.00%。结论:采用基于决策树的算法可以筛选出CHD&DM患者气阴两虚证的相关理化指标,并建立CHD&DM患者气阴两虚证的理化指标诊断模式,具有一定的实用价值,应进一步进行深入研究。

Objective： To sieve physicochemical indexes most related with dual deficiency of qi and yin pattern by data mining and effective data processing in patients with coronary heart disease（ CHD） combining with diabetes mellitus（ DM）in order to establish the diagnostic mode between physicochemical indexes and syndrome factors. Methods： Clinical basic data,information of the four diagnostic methods in traditional Chinese medicine,and physiochemical indexes from 100 CHD patients with unstable angina combining with DM were collected and comprehensively analyzed. On the basis of t test,nonparametric analysis,and Spearmancorrelation analysis,the correlation between dual deficiency of qi and yin pattern and physicochemical indexes was explored by decision tree,one of data mining methods. Results： By decision tree,correlation between dual deficiency of qi and yin pattern and clinical physicochemical indexes in CHD patients with angina pectoris combining with DM was processed with data mining and analysis. The decision tree mode of dual deficiency of qi and yin patter from 6 physicochemical indexes including ejection fraction,low density lipoprotein cholesterol（ LDLC）,2-h plasma glucose（ 2 hp G）,time limit of QRS,serum progesterone（ P）,and triglyceride（ TG） was preliminarily obtained. There were 7 recognition pathways of dual deficiency of qi and yin pattern and non-dual deficiency of qi and yin pattern,and the accuracy rate of detection reached 77. 00%. Conclusion： The related physicochemical indexes can be sieved by decision tree in CHD patients with angina pectoris combining with DM in dual deficiency of qi and yin pattern. The diagnostic mode of dual deficiency of qi and yin pattern from physicochemical indexes is established in CHD patients with angina pectoris combining with DM,which has a certain practical value and should be further studied.