文献类型：期刊文献

中文题名：基于递归定量分析的冠心病中医脉象特征提取与分析

英文题名：Study of traditional Chinese medicine pulse signals in patients with coronary heart disease based on recurrence quantification analysis

作者：郭睿[1];王忆勤[1];颜建军[2];燕海霞[1];杨育慈[1]

机构：[1]上海中医药大学四诊信息综合实验室,上海201203;[2]华东理工大学机械电子工程中心,上海200237

年份：2011

卷号：9

期号：11

起止页码：1226

中文期刊名：中西医结合学报

外文期刊名：Journal of Chinese Integrative Medicine

收录：MEDLINE(收录号：22088589);;Scopus;PubMed

基金：上海市重点学科(第3期)中医诊断学经费资助项目(No.S30302);上海市教育委员会科研创新项目(No.11YZ71)

语种：中文

中文关键词：冠心病;脉象;递归定量分析;脉象图;临床试验

外文关键词：coronary disease; pulse presentations; recurrence quantification analysis; sphygmogram; clinica trial

摘要：目的:基于非线性动力学方法研究冠心病患者的脉象,寻找能区分冠心病和正常人脉象的非线性动力学参数。方法:基于非线性动力学理论,对63例冠心病患者和61例正常人的脉搏信号的递归图进行递归定量分析(recurrence quantification analysis,RQA),计算两组脉搏信号的RQA参数,这些参数包括递归率(recurrence rate,RR)、确定性(determinism,DET)、对角线长度的均值(averaged diagonal length,L)、递归熵(entropy of diagonal length,ENTR)、最长对角线(length of longest diagonal line,L_(max))、层状度(laminarity,LAM),竖直/水平线段长度均值(trapping time,TT)和最长竖直/水平线段长度(length of longest vertical line,V_(max));并用秩和检验统计两组脉搏信号的RQA参数之间的差异,用箱线图(BoxPlot)观察有显著性差异的RQA参数的数值分布;绘制有显著性差异的RQA参数的接受者工作特征(receiver operatingcharacteristic,ROC)曲线,对RQA参数进行诊断性试验,以评估脉搏信号的RQA参数区分冠心病组和正常组的诊断价值。结果:秩和检验统计结果表明冠心病组脉搏信号的RQA参数中RR、DET、L、ENTR、LAM、TT和V_(max)的平均秩次显著大于正常组,它们的ROC曲线下的面积依次为1.000、0.898、0.653、0.673、0.885、0.898、0.986和0.994。结论:与正常组相比,冠心病组的脉搏信号表现出较高的规律性、确定性和稳定性;脉象的非线性动力学RQA参数RR、TT、V_(max)DET和LAM对区分冠心病组和正常组具有较高的诊断价值。这为脉诊的无损伤检测提供了新的手段。
Objective： By using recurrence quantification analysis （RQA） to analyze traditional Chinese medicine pulse signals of patients with coronary heart disease （CHD）, this study aims to find nonlinear dynamic parameters of pulses to distinguish patients with CHD from normal subjects. Methods： First, pulse signals were collected using ZBOX-I pulse digitization gathering analyzer from October 2007 to June 2008. RQA was used to analyze RQA parameters of pulses of 63 patients with CHD and 61 normal subjects. RQA parameters included recurrence rate （RR）, determinism （DET）, averaged diagonal length （L）, entropy of diagonal length （ENTR）, length of longest diagonal line （Lmax）, laminarity （LAM）, trapping time （TT） and length of longest vertical line （Vmax）. Then, rank-sum test and BoxPIot were employed to find significant difference and distribution of RQA parameters. Lastly, receiver operating characteristic （ROC） curves were used to assess the diagnostic value of the measurements with significant difference. Results： There were significant differences in RQA parameters of pulse signals between the two groups, including RR, DET, L, ENTR, LAM, TT and V and their areas under the ROC curves were 1.000, 0. 898,0. 653,0. 673,0. 885, 0. 898, 0. 986 and 0. 994, respectively. Conclusion： Compared with the normal subjects, the pulse signals of the patients with CHD are presented with more certainty, regularity and stability. RQA measurements of RR, TT, V DET and LAM show good diagnostic value according to their ROC curves.