An ECG denoising method combining variational modal decomposition and wavelet soft threshold

Abstract

AbstractCardio‐cerebrovascular disease is one of the malignant diseases with the highest morbidity and mortality in the world. Medical research shows that the factors that induce heart blood cerebrovascular disease are diverse, and electrocardiogram (ECG) has important reference significance for the diagnosis and pathological research of cardiovascular and cerebrovascular diseases. This paper proposes a method of denoising ECG signals by combining variational mode decomposition with wavelet soft‐threshold. First perform variational modal decomposition on the ECG signal to obtain several modal components, and then the decomposed modals components are denoised by wavelet threshold, and finally the signal reconstruction of the denoised modal components is carried out through Matlab simulation experiment analysis. Compared with the soft‐threshold denoising method of ensemble empirical mode decomposition (EEMD), the soft‐threshold denoising method of complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), the variational modal decomposition wavelet threshold denoising algorithm combined with ECG signal denoising has better denoising effect. This method is used to denoise the ECG signal from the MIT‐BIH database of the Massachusetts Institute of Technology. Under the condition of ensuring the smoothness of the ECG signal image, the signal‐to‐noise ratio value is 27.1121 and the mean square error value is 0.0535, which has achieved good results.