摘要
目的:去除在心电信号采集过程中混入的肌电干扰、工频干扰、基线漂移等噪声信号,避免噪声对心电信号特征点的识别和提取造成误判和漏判。方法:首先利用coif4小波对心电信号按Mallat算法进行分解,然后采用软、硬阈值折衷与小波重构的算法进行去噪。结果:采用MIT/BIH Arrhythmia Database中的心电信号进行仿真、验证,有效去除了三种常见的噪声信号。结论:本方法实时性好,为临床分析与诊断奠定了基础。
Objective： In oder to avoid the wrong identification from the characteristics of ECG signals, to remove the electromyo- graphical interference （EMG）, powerline interference and baseline-drift existing in the course of collecting ECG signal data were removed importantly. Methods： Firstly, based on the Mallat algorithm, the ECG signals was decomposed through the coif4 wavelet. Secondly, the denoising algorithm of tradeoff of soft-and-hard thresholding and the wavelet reconstruction algorithm were used to remove the main noise. Results： The MIT/BIH arrhythmia database was used, which proved that the algorithm could effectively remove the noise. Conclusions： This algorithm is suitable for real-time analysis, and lays a good basis for the clinical analysis and diagnosis.
出处
《现代生物医学进展》
CAS
2007年第10期1566-1568,共3页
Progress in Modern Biomedicine
关键词
小波变换
心电信号
工频干扰
肌电干扰
基线漂移
Wavelet
ECG signal processing
Powerline interference
Electromyographical interference
Baseline-drift
作者简介
赵晴，（1983-），女，硕士研究生，主要研究方向：生物医学信号处理。Email：qinger802@126．com，地址：山东省济南市文化东路88号，250014。
