Denoise of ECG Based on Weighted Sum of Intrinsic Mode Functions

Denoise of ECG using weighted sum of intrinsic mode functions (IMFs) with weights determined by the distance metric between each IMF and the noise is proposed in this paper. Traditionally, approaches of the like are achieved by subtracting some IMFs directly from the noisy ECG. However, the determination of IMFs as the subtrahends to the noisy ECG is subject to trial and error. The new approach avoids the problem of trial and error and hence is much easier to apply. Several methods for computing the distance metrics are compared in this paper. Computer simulation results show that the method performs well compared to the ones by directly subtracting some IMFs from the noisy ECG.