基于心脏电-机械活动变异性的冠心病分布熵特征研究

项目名称： 基于心脏电-机械活动变异性的冠心病分布熵特征研究

项目编号： No.61471223

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 无线电电子学、电信技术

项目作者： 刘常春

作者单位： 山东大学

项目金额： 80万元

中文摘要： 从心电和心音信号提取冠心病特征具有重大的科学价值和临床意义。本项目拟发展一种单变量和多变量分布熵算法，并用于研究心率变异、舒张间期变异、电机械延迟时间变异特征和三者间的耦合特征，为无创无损检测冠心病提供创新的思路和技术支持。 研究内容：1)发展单变量分布熵算法：在序列相空间中，基于向量间距离分布的全局概率密度函数，构建分布熵理论和算法。2)发展多变量分布熵算法：提出多通道序列的多变量相空间重构方法，构建多变量分布熵理论和算法。3)研究对象：健康对照组、I级-IV级4个冠心病组；由同步采集的心电和心音信号，构建电机械延迟时间、舒张间期和心动周期序列。4)分别计算各研究对象的上述3个序列的单变量分布熵，研究各组的单变量分布熵特征。5)分别计算各组的3个序列耦合的多变量分布熵，研究其多变量分布熵特征。6)融合4)、5)所得的4个分布熵参数，建立冠心病无创无损检测指标。

中文关键词： 心脏电-机械活动；变异性；分布熵；冠心病特征；辅助诊断

英文摘要： It is of great significance and clinical prospect to seek signs of coronary heart disease (CHD) from ECG and heart sound. This proposal aims to develop both a uni- and a multivariate distribution entropy measure, and to investigate the corresponding univariate characteristics of heart rate variability (HRV), diastolic period variability (DPV), and electro-mechanical delay variability (EMDV), as well as the multivariate coupling characteristics of HRV-DPV-EMDV. Novel ideas and techniques will be provided through this project for the noninvasive and non-destructive detection of CHD. Main works of this project are listed as follows. 1) To develop the univariate distribution entropy algorithm. A key technique for the development lies on the global quantification of the state space. It is intended to be achieved by the probability density function of distances across vectors. 2) To develop the multivariate distribution entropy measure. A multivariate state space reconstruction method will be developed and this multivariate measure will be constructed accordingly. 3) Subjects include the normal control group and 4 CHD groups with different degrees of coronary artery stenosis. The EMDV, DPV, and HRV series will be constructed from simultaneously recorded ECG and heart sound signal. 4) To investigate the corresponding univariate characteristics of the 5 groups through the calculation of the univariate distribution entropy measures of HRV, DPV, and EMDV. 5) To seek the multivariate coupling characteristics of all groups by the multivariate distribution entropy of HRV-DPV-EMDV. 6) To develop a novel index for detecting CHD without invasion and destruction by the fusion of the aforementioned 4 distribution entropy results obtained in 4) and 5).

英文关键词： Cardiac electrical and mechanical activity;Variability;Distribution entropy;Features of coronary heart disease;Aided diagnosis