心率变异的非线性特征参数估算  

Estimations of Nonlinear Characteristic Parameters for Human HRV Time Series

作　　者：王黎[1] 于涛[1] 韩清凯[1] 闻邦椿[1] 

机构地区：[1]东北大学机械工程与自动化学院,沈阳110004

出　　处：《非线性动力学学报》2005年第1期8-16,共9页

摘　　要：心率变异（Heart Rate Variability,HRV）是指人的心脏节律的微小变动量，与人的健康状态和精神状态直接相关，具有明显的非线性特征。在本文中，对HRV时间序列的几个非线性特征参数进行估算．从而对心脏健康状态（心率正常）与非健康状态（心率变异）HRV之间的差别进行比较。首先利用小波变换技术对心电信号（ECG）数据进行R波的准确定位，经过重采样得到HRV序列。关联维的计算结果表明，健康状态和非健康状态HRV时间序列具有不同的分形结构，在相空间重构的基础上对HRV进行最大李雅普诺夫指数的估算。结果表明，健康状态和非健康状态HRV时间序列的最大李雅普诺夫指数均为正值，但处于心率不齐状态的节律的混沌程度明显低于健康状态，健康状态HRV的复杂度要高于非健康状态HRV的复杂度，近似嫡和复杂度的分析结果基本相似，健康状态HRV的近似熵要高于非健康状态HRV的近似熵。利用这些非线性特征参数对健康状态和非健康状态的HRV进行比较分析，可以为诊断提供依据。HRV （Heart Rate Variability）, referred to the weak differences among every normal heart beating period, is directly relative to the human body healthy state. It appears obviously nonlinear characteristics. In the paper, some nonlinear parameters of HRV are calculated and compared for that of both healthy （normal heart rhythm） and unhealthy （arrhythmia） people. Firstly, The R wave peaks of ECG are located precisely with wavelet transform technique, and the HRV time series are achieved after re-sampling. The calculated correlative dimensions of HRV show that HRVs are fractal in geometry. The largest Lyapunov exponents of HRV are estimated on the base of phase space re-construction. The obtained largest Lyapunov exponents, both of healthy and of unhealthy, are all positive and the later is less than the former in chaos. The approximate entropy of healthy HRV is higher than that of unhealthy. The complexity values show alike as the approximate entropy for both cases, and the healthy values are higher than the unhealthy case. The comparisons of these nonlinear parameters of HRV are good references in diagnosing of ECG signals for the arrhythmia and may be helpful in researching of living sciences in future.

关 键 词：HRV 相空间重构 关联维 李雅普诺夫指数 复杂度 近似熵 

分 类 号：O322[理学—一般力学与力学基础] TH113.1[理学—力学]