一种改进的心电信号预处理方法研究

Research on improved ECG pre-processing method

心电信号是典型的强噪声下的非平稳微弱信号,减小噪声的干扰对心电信号的分析有着十分重要的意义,因此,有效的滤波方法一直是该领域学者关注的热点问题。本文在基于小波变换心电信号分析研究基础上,针对小波去噪时分解只作用于低频部分,从而忽略了高频区域中一部分有用信号的问题,提出了一种采用改进小波包理论实现心电信号去噪的方法,利用小波包在消除信号噪声方面具有更为精确的局部分析能力的特点,采用了‘db4’小波和"最优基"选择的方法,对心电信号进行消噪。以MIT-BIH心电数据库中心律失常数据仿真实验,得到了较理想的去噪效果。对比该方法与小波滤波去噪,发现基于小波包的心电信号去噪具有更优良的去噪性能。

Electrocardiogram （ECG） is a typical weak signal with un-smooth characteristics and heavy noises.It is important of researching an effective method to weak the influence of noises.Aimed at the problem of removing the noises using WT method,that is only the low frequency domain of ECG signal was decomposed.An improved ECG pre-processing method was researched in this paper.When using the Wave-Package Transform （WPT） method to weaken the noises,only the noise in the high frequency domain is decomposed,but some useful information maybe eliminated. So we choose the wavelet ＇db4＇and the best radix to remove the noise of the signal. In order to compare the two methods, the simulation was completed with the MIT/BIH Database. The better simulation result was achieved with wave-package transform （WPT） than wavelet transform （WT）.