An effective electrocardiogram segments denoising method combined with ensemble empirical mode decomposition, empirical mode decomposition, and wavelet packet

Electrocardiogram (ECG) is the most extensively applied diagnostic approach for heart diseases. However, an ECG signal a weak bioelectrical and easily disturbed by baseline wander, powerline interference, muscle artefacts, which make detection of diseases more difficult. Therefore, it very important to denoise contaminated in practical application. In this article, effective segments denoising method combining ensemble empirical mode decomposition (EEMD), (EMD), wavelet packet (WP) designed. The decomposed using EEMD first time, then highest frequency component EMD second high components obtained from time are reconstructed WP third time. Finally, processed fused obtain denoised signal. Furthermore, signal-to-noise ratio (SNR), mean square error (MSE), root (RMSE), normalised cross correlation coefficient (R) used evaluate noise reduction algorithm. SNR, MSE, RMSE, R 5.7427, 0.0071, 0.0551, 0.9050 China Physiological Signal Challenge 2018 dataset. Compared with others methods, experimental results not only exhibit that SNR effectively improved, but also show details fully retained, laying solid foundation automatic segments.