基于CEEMD的心音信号小波包去噪算法研究

针对传统心音去噪方法易将其部分高频有用信息作为噪声滤除而造成滤波后的心音信号失真及信息丢失的问题,提出了一种基于互补总体经验模态分解(CEEMD)的小波包变换去噪算法。首先通过互补总体经验模态分解将心音信号分解为从高频到低频的不同固有模态函数分量(IMFs),并利用自相关函数客观界定信号的模态分量范围;然后对噪声主导模态分量和混叠模态分量采用小波包变换进行滤波提取有用信息后,与剩余固有模态分量进行重构得到去噪后的信号。实验结果表明,改进的算法不仅可以去除心音中的噪声成分,明显改善心音信号的信噪比和均方根误差,而且能够有效保留信号的高频有用信息,且在不同噪声水平下的去噪性能均优于传统算法,鲁棒性较好。

Wavelet packet de-noising algorithm for heart sound signals based on CEEMD

Here, aiming at problems of traditional heart sound signals’ de-noising method being easy to eliminate parts of high frequency useful information and cause distortion of heart sound signals and loss of information, a wavelet packet de-noising algorithm based on the complementary ensemble empirical mode decomposition (CEEMD) was proposed. Firstly, heart sound signals were decomposed into different intrinsic mode functions (IMFs) with CEEMD. Autocorrelation function was used to objectively define the range of modal components of a signal. Then, the useful information was extracted from noise dominant modal components and aliasing ones using the wavelet packet transformation, and it was used for reconstruction of the de-noised signal together with residual IMFs. The results showed that the proposed method can be used not only to eliminate noise components in heart sounds, and improve heart sounds’ ratio of signal to noise and the root mean square error, but also to effectively retain heart sound signals’ high-frequency useful information;compared with traditional algorithms, it has better de-noising performance and robustness under different noise levels.