基于小波变换的改进阈值法在心电信号去噪中的应用

Application of modified threshold method based on wavelet transform in ECG signal de-noising

目的为了克服传统小波硬阈值函数的不连续性和软阈值函数有偏差的缺点,并消除采用固定阈值带来的偏差。方法改变小波阈值函数部分参数得到一种新的小波阈值函数,新的小波阈值函数可以看成是硬阈值函数和软阈值函数的线性组合,既克服了软硬阈值函数的缺点,同时又能在它们之间进行灵活选择,小波阈值的选取采用了一种自适应阈值。结果改进阈值法能有效去除心电信号中常见的工频干扰、肌电干扰和基线漂移3种噪声。结论与软硬阈值法相比,无论是从视觉效果上还是信噪比上都有较大的改善。

Obj ective To avoid the disadvantages of traditional wavelet threshold function, and eliminate the error caused by fixed threshold. Methods New wavelet threshold function was obtained by changing some parameters of traditional wavelet threshold function. The new wavelet threshold function were taken as linear combination of hard and soft threshold functions. At the same time, the modified threshold method avoid the shortcomings of soft and hard threshold function, and switched flexibly between them, and the wavelet threshold was chosen via an adaptive threshold. Results The modified threshold method was effectively remove three kinds of common noises such as frequency interference, electromyography interference and baseline drift. Conclusion It is demonstrated that the modified threshold method greatly improves in both visual effect and signal to noise ratio compared with hard and soft threshold method.