一种新的基于Slantlet变换的心电信号消噪算法

Novel Denoising Algorithm for Electrocardiogram Signals Based on Slantlet Transform

心电信号在采集过程中常常掺杂着各种噪声。针对心电信号多突发性与多间断性的特点,采用一种改进的正交离散小波变换——Slantlet变换对其进行消噪处理。由于传统的小波阈值消噪算法在信号中的奇异点处会产生伪Gibbs现象,引入平移不变方法使伪Gibbs现象得到有效的抑制,并利用Slantlet基函数的分片线性的优势,提出一种新的基于信号重采样与圆周平移不变变换的消噪方法。利用美国麻省理工学院的PhysioBank生理信号数据库对以上方法进行验证,实验结果表明该算法能有效地消除噪声并较好地保持心电信号的几何特性。

The electrocardiogram signals （ECG） may be mixed with various kinds of noises while being gathered and recorded. Since ECG signals possess many jumps and abrupt places, the proposed denoising algorithm employs an improved version of orthogonal discrete wavelet transform （DWT） called Slantlet Transform. Traditional wavelet threshold methods may cause Pseudo-Gibbs phenomena on the singular points of the signals. The denoising algorithm of wavelet threshold based on Translation-Invariant was applied to eliminate the phenomenon; furthermore, with the benefit of Slantlet piecewise basis, a novel algorithm based on signal resampling and Cycle-Spinning Translation-lnvariant was proposed. The PhysioBank database of MIT was used to validate the algorithm. Experimental results show that the proposed algorithm can effectively remove noise and well preserve the geometric characteristics of the electrocardiogram signals.