Electrocardiography(ECG) signal suffers from noise interference during the recording and transmission, which would result in morphological structure distortion and affects proper diagnosis of diseases related to the cardiac system. For the purpose of filtering the noise to obtain the clean signal, in this paper, the recently developed complete ensemble empirical mode decomposition with adaptive noise algorithm(CEEMDAN) with morphological structure of ECG signal is proposed for ECG enhancement. The experimental results indicate that the proposed CEEMDAN based method could be capable of removing both high frequency interference and low frequency baseline wander, which provides competitive performances compared to other signal de-nosing approaches explored in this paper.

心电图（ECG）信号在记录和传输过程中会受到噪声干扰，这会导致形态结构变形并影响与心脏系统相关疾病的正确诊断。为了过滤噪声以获得纯净的信号，本文提出了一种新的基于ECG信号形态结构的自适应噪声算法（CEEMDAN）完全集成的经验模式分解方法，用于ECG增强。实验结果表明，基于CEEMDAN的方法可以消除高频干扰和低频基线漂移，与本文探讨的其他信号去噪方法相比，具有竞争优势。