A review of noise removal techniques in ECG signals

In this paper, three improved methods are explored to address the shortcomings of existing ECG signal denoising methods. The first method is based on the empirical identification of the intrinsic mode function (IMF) component of the QRS eigenwave and the reconstruction of the inaccurate identification of the ECG signal, using the statistical properties of the EMD and IMF components to denoise the ECG signal. The second method is out to use a combination of variational modal decomposition and wavelet thresholding to denoise the ECG signal for inotropic interference. The third method is to propose a denoising algorithm based on improved wavelet thresholding-CEEMDAN to address the shortcomings of the wavelet thresholding method. All three schemes have evaluation metrics and criteria and have been tested and proven to be substantially better than traditional methods to a certain extent.