基于等概率粗粒化的复杂度算法及其应用  

Complexity Algorithm Based on Equiprobable Coarse Graining and Its Application

作　　者：张佃中[1] 谭小红[2] 王智[1] 刘昭前[3] 

机构地区：[1]中南大学数学科学与计算技术学院,湖南长沙410083 [2]中南大学物理科学与技术学院,湖南长沙410083 [3]中南大学临床药理研究所,湖南长沙410078

出　　处：《系统仿真学报》2008年第15期4096-4098,4103,共4页Journal of System Simulation

基　　金：国家自然科学基金(30572230)

摘　　要：为探究Lempel-Ziv复杂度(LZC)在心率变异(HRV)分析中的稳定性,提出了等概率粗粒化重构出符号序列计算LZC的新方法和具体算法.以年轻(21-34岁)与年老(68-85岁)组健康受试者的HRV信号为实验数据,选取不同的粗粒化段数,用常规和等概率方法分别计算LZC.再算出不同分段数LZC序列之间的相关系数.结果表明,用新方法得出的相关系数都大于0.96,年老组的LZC指标明显降低(P<0.001),这些结果不受粗粒化段数的影响.新方法算得的LZC是度量HRV稳定而有效的指标。In order to explore the stability of Lempel-ziv complexity （LZC） to analyze heart rate variability （HRV）, a new method and its detailed algorithm were proposed, in which LZC was calculated using equiprobable coarse graining to reconstruct the symbol series. HRV signals were used as experimental data, which were obtained from young （21-34 yr） and elderly （68-81 yr） groups of healthy subjects. Various segmented number was chosen, LZC was worked out using conventional and equiprobable coarse graining method respectively, and the correlation coefficient of the complexity series of various segmented number was worked out. The results show that all the correlation coefficients are 0.96 plus and LZC index is obviously lower in elderly group （P〈0.001） with the new method, and they are unaffected by the segments of coarse graining. LZC is a steady and effective index to measure HRV with the new method.

关 键 词：Lempel-Ziv复杂度 粗粒化 等概率 心率变异 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]