An inverse Bayesian scheme for the denoising of ECG signals

The ECG filtering problem is a widely explored research topic. In this paper we propose a novel methodology taking into account some “a priori” information on the signal components that gives significant improvements in the denoising task. Our approach is a modification of recursive filtering (RF) by considering a probabilistic Bayesian framework that is adopted to remove selectively artifacts from the signals without the loss of any crucial medical information on the acquired data. The proposed scheme is very efficient from the computational point of view and this feature makes it ready as a useful tool for application in mobile devices equipped with specific sensors. Finally, numerical experiments have been carried out on real data to assess the advantages of the algorithm.