三种时频分析方法在心音信号分析中的应用  

A Study of Using Three Time-frequency Methods in theAnalysis of Heart Sound Signals

作　　者：张建[1] 沈民奋[1] 宋骥[1] 

机构地区：[1]汕头大学电子系,汕头515063

出　　处：《汕头大学学报（自然科学版）》2003年第2期62-69,共8页Journal of Shantou University：Natural Science Edition

摘　　要：文章介绍了一种新的时频分析方法—— Hilbert谱分析 ,并将其应用到心音信号的时频分析中 ,分析对比了短时 Fourier变换 ,小波分析和 Hilbert谱分析这三种方法在处理心音信号时的性能差异 .传统的基于 Fourier的方法不能有效的分析非平稳信号 .文章分析证明了这种新方法在分析像心音信号这种非平稳信号的有效性 ,它能对信号的时频分布做出比短时 Fourier变换和小波分析更为精确的描述 .用这种方法 ,我们能提取心音信号中频率变化的信息 .We will introduce a new time-frequency analysis approach,Hilbert spectral analysis method,and its application to heart sounds signals.The differences in the performance of the short Fourier Transform,continuous wavelet transform and Hilbert spectrum in time-frequency analysis of heart sound signals are studied and compared in this paper.Conventional Fourier-based methods are insufficient for analyzing non-stationary data.This paper illustrates that this method is adequate for non-stationary data,such as heart sounds signals.This new method gives a more precise definition of particular events in time-frequency space than the short Fourier Transform and wavelet analysis.We can use the method to resolve changes in the frequency content of the heart sounds signals.

关 键 词：心音信号 时频分析 非乎稳信号 短时Fourier变换 小波分析 Hilbert谱分析 

分 类 号：R540.44[医药卫生—心血管疾病] TN911.6[医药卫生—内科学]