心音信号噪声消除的小波变换方法  

Denoising Methods Of Heart Sound Signal Based on Wavelet Transform

作　　者：张佃中[1] 谭小红[2] 

机构地区：[1]中南大学数学科学与计算技术学院生物数学与生物信息部,湖南长沙410078 [2]中南大学物理科学与技术学院,湖南长沙410078

出　　处：《生物数学学报》2004年第4期477-480,共4页Journal of Biomathematics

摘　　要：心音信号幅值小，干扰多，采用常规的时、频域滤波方法往往不能收到良好的效果。本文根据信号和干扰在小波变换下的不同变化特性，利用二进小波变换的模极大值识别出心音信号中的干扰噪声的位置，剔除其相应的小波变换系数后，再通过小波逆变换重构出心音信号，并根据心音信号的特点选取了适当的母小波和分解尺度，给出了利用小波方法去噪前后的实际结果，结果表明，小波变换方法可有效地消除心音信号中的噪声干扰。Heart sound signal is weak and disturbance is wide-spread, conventional timefrequency methods are of little effcet. In this paper, based on the behavior of different changing trends between signal and the disturbance in the wavelet transform, the position of the disturbance signals are deteted based on the dyadic wavelet transform modulus maxima. Heart sound signal is reconstructed by wavelet inverse transform which used wavelet transform coefficients eliminating the ones of the disturbance signal. A proper mother wavelet and decomposing scales are chosen based on characteristics of the heart sound signal. Along with comparison plots of original signal and denoised signal, the results show that the heart sound signal are denoised in the wavelet transform methos effectively.

关 键 词：心音信号 小波变换 尺度 模极大值 消噪 

分 类 号：O29[理学—应用数学] R311[理学—数学]