临床心电信号工频干扰小波去噪方法对比分析  

Powerline Interference Denoising Algorithm Analysis of ECG Signals Based on Wavelet Transform

作　　者：张雪[1] 王海燕[1] 李保军[1] 

机构地区：[1]西北工业大学航海学院,陕西西安710072

出　　处：《计算机测量与控制》2010年第4期902-905,共4页Computer Measurement &Control

摘　　要：针对临床心电信号存在的工频干扰,基于小波变换方法,从实用性角度对比分析了三种去噪方法;频域强制去噪法:先对小波系数进行傅立叶分解,将受噪声干扰严重的系数置零,再进行小波重构;此法得到的信号较光滑,但有可能丢失有用信息;能量阈值去噪法:根据提前设定的能量恢复指数,得到预定阈值作为门限,保留一定的细节系数和全部近似系数重构小波;此法简化并优化了传统阈值的选取,较好保留了信号中的突变部分;默认阈值去噪法:其可信度较低,但仍可满足非高精度数据处理的要求;在实际临床应用中,可根据实际情况和处理精度选取合适的去噪方法,为进一步研究临床心电信号预处理奠定基础。Focus on the powerline interference of ECG signals, this paper analyzed three ways of denoising algorithm based on Wavelet. Mandatory denoising in frequency domain was to make Fourier decomposition of the wavelet coefficients, and then let the infected coefficients to be zeros. The denoising way based on energy threshold function was giving wavelet thresholding and reconstructing technique and the performance of this method was well. Default threshold denoising way with a lower reliability can be used for low-accuracy requirement. Combined with MATLAB simulation and clinical data, it was proved that three methods can selectively applied for powerline interference denoising and are suitable for real-time analysis, which lay good basis for the clinical analysis and diagnosis.

关 键 词：小波去噪 强制去噪 能量阈值去噪 默认阈值去噪 

分 类 号：TN911.72[电子电信—通信与信息系统]