【摘要】 针对心电信号(ECG)采集过程中常受到白噪声干扰的情况,提出了一种基于多种基函数的形态分量分析(MCA)去噪方法。MCA利用信号组成成分的形态差异性通过超完备字典对其进行稀释表示,使得各形态成分得到有效分离。根据ECG特征波形的形态差异性,选用离散余弦变换字典来稀疏表示心电信号中的平滑成分(如P波),选用非抽样小波变换字典来稀疏表示心电信号中的突变成分(如QRS波群),同时滤除高频噪声成分。对MIT-BIH心率失常数据库中的样本进行仿真实验。结果表明,该方法去噪效果优于小波消噪法,不仅可以有效地抑制心电信号中的白噪声干扰,还能较好地保留ECG特征波形。更多还原

【Abstract】 In the process of ECG signal data acquisition,it was easy to involve white noise. Based on morphological component analysis( MCA),a novel method for ECG de-noising was proposed. The method uses a variety of forms of decomposition dictionaries. MCA is a new method which takes advantage of the sparse representation of structured data in overcomplete dictionaries to separate features in the data based on the diversity of their morphology. According to the morphological differences in ECG waveform characteristics,Discrete Cosine transform( DCT) dictionary was used to sparsely represent ECG smoothing component( such as P wave), and undecimated Wavelet Transform( UWT)dictionary was used to sparsely represent mutations component( such as QRS wave),then the white noise component could be filtered out. Through using the MIT-BIH database of ECG signal for simulation,conclusion shows that the method proposed in this paper performs better than wavelet de-noising method significantly,not only in suppressing white noise effectively,but also in preserving the primary characteristics waveform of ECG.更多还原