基于小波变换的动态心电信号伪差识别

Artifacts Discriminating of Dynamic Electrocardiogram Signals Based on Wavelet Transform

由突变干扰引起的伪差会严重阻碍动态心电信号的自动分析和正确诊断,常规的小波消噪方法无法消除这类伪差干扰。提出一种新的识别心电信号突变干扰的方法,即利用小波多尺度分解并结合阈值判定算法对突变干扰这类伪差进行自动识别。采用实测的30组动态心电信号对该算法进行测试,实验结果表明,该算法能快速、有效地识别动态心电信号中的突变干扰,正确检出率达到96.4%,为正确诊断动态心电信号提供保障。

Artifacts which are caused by the mutation noise disturbance often hinder Dynamic electrocardiogram（DCG） automated analysis and the correct diagnosis. The conventional wavelet denoising method can not eliminate this kind of artifacts. This paper proposes a new method for discriminating the artifacts which is based on multi-scale wavelets decomposition and threshold determination algorithm. Actual 30 group DCG data are used to test the algorithm. Experimental result shows that it is possible to quickly and effectively discriminate the artifacts, the reliability of this algorithm in accurately detecting the artifacts location is 96.4%, the new method is conducive for ensuring clinicians＇ correct judgment of DCG signal.