Power Line Interference Suppression for ECG Signal Recovery

The electrocardiogram (ECG) recovery is very important for clinical diagnosis, especially in the presence of power line interference (PLI). In this work, to suppress the PLI, it is modeled as a linear combination of sinusoidal signals that have a sparse representation in the frequency domain. To accurately reconstruct the ECG, the time domain as a sparse domain for ECG signal is exploited. Based on the sparse representations, a joint optimization estimation problem is developed that allows one to simultaneously perform the ECG recovery and PLI suppression. In order to solve the optimization problem, two efficient schemes based on the greedy algorithms such as orthogonal matching pursuit (OMP) and compressive sampling matching pursuit (CoSaMP) are utilized. Finally, numerical studies demonstrate that the JCoSa.MP estimation algorithm outperforms the state-of-theart approaches.