中文摘要：

目的：基于样本熵分析冠心病组和正常组脉诊信号并分析其差异。方法：计算61例冠心病患者和61例正常人的脉诊信号的窗口样本熵的参数,将其转换成秩次后进行方差分析,最后绘制这些参数的接受者工作曲线（ROC）来评估它们区分冠心病组和正常组的诊断价值。结果：冠心病组的基于窗口样本熵计算的参数显著小于正常组的,说明冠心病脉诊信号的复杂度比正常组的低,并且冠心病脉诊信号的形态变异性比正常组小。脉象窗口样本熵均值参数的ROC曲线下面积为0.955。结论：脉象样本熵参数为冠心病脉象的病理状态提供了一种较好的表征方法;窗口样本熵均值能较好的区分冠心病脉象和正常人脉象。

英文摘要：

Objective： Sample entropy analysis was used to investigate the differences of the TCM pulse signal of the patients with coronary heart disease （CHD） and the normal subjects. Methods： Firstly parameters of window sample entropy of pulses from 61 patients with CHD and 61 normal subjects were calculated, and then the parameters were converted to rank for Variance-test. Lastly, receiver operating characteristic （ROC） curves were used to assess the diagnostic value of parameters for differentiating the CHD from the normal subjects. Results： The parameters of pulses of the patients with CHD were significantly lower than those of the normal subjects, which meaned that the pulse signal of patients with CHD were lower of complexity than those of normal ones. And the pulses with CHD had small morphological variability. The area under ROC curve of the parameter, the mean of window sample entropy of pulses, was 0.955. Conclusion： The windows sample entropy provided a good index for pathological state of pulse with CHD. The parameter, the mean of window sample entropy of pulses, showed good diagnostic value for differentiating the CHD from the normal subjects according to its ROC curves.