Research And Experimental Verification Of Quantitative Adaptive ECG Denoising Method Based On CEEMD

Electrocardiogram(ECG)signal,one of the earliest bioelectric signals applied in clinical medical research,is an important physiological signal in the human body,which can comprehensively reflect the health status of the human body and has important significance for the diagnosis and prevention of cardiovascular diseases.However,it is often accompanied by noise interference when using medical equipment to collect ECG signals.Therefore,how to extract pure ECG signals in practical applications is an urgent problem to be solved.In order to better solve the above problem and enable the collected ECG signals to be used more accurately,a new quantitative adaptive ECG denoising method based on CEEMD was studied in this paper.This method is based on the in-depth study of EMD and its optimization method.Firstly,the original signal is decomposed effectively by CEEMD,and a series of IMF components are finally obtained.For IMF components,this paper adopts the principle of sub-domain to deal with them respectively.Correlation analysis is used to calculate the correlation between the original signal and each IMF component,and then the cutoff point between the noise domain and the signal domain is obtained.Then the effective modal functions are selected in the signal domain according to the energy priority and the noise domain according to the low frequency priority.Finally,the corresponding correlation threshold is set according to the mutual information calculation,and the selected effective modal functions are reconstructed through the correlation threshold setting to obtain the final de-noising signal.In the new denoising method of ECG proposed in this paper,each parameter can be calculated quantitatively and adaptively,which avoids the need of prior knowledge and greatly improves the accuracy and adaptability of the denoising method.In addition,in view of the obvious interference of baseline drift in some ECG signals collected,it is proposed that morphological filtering can be used for preprocessing to correct the baseline drift of signals.Then a new de-noising method is used to remove the high frequency noise interference.Finally,both of them have been verified by digital simulation experiment and experiment of measured signal.The digital simulation experiments of Blocks,Doppler and Heavy sine were used to verify the new ECG denoising method,and the wavelet threshold denoising method was also compared.The results of the two methods were analyzed and verified.In order to better prove the practicality of the new ECG de-noising method,the experimental verification was carried out by using the measured signal and the clinical medical signal in the MIT-BIH database,the experimental exploration of the new ECG de-noising method based on CEEMDAN was also carried out,and the results were analyzed and studied.The quantitative adaptive ECG de-noising method based on CEEMD developed in this project has shown good de-noising effect through digital simulation experiment and experimental verification of measured signals.In the end,the de-noising signals have higher signal noise ratio and higher correlation with the original signal.Therefore,in practical application,this method can obtain relatively pure ECG signals,providing a reliable basis for the diagnosis and prevention of cardiovascular diseases.