Extraction of ECG from single channel EMG signal using constrained singular spectrum analysis

A simple and extremely effective method for extraction of electrocardiography (ECG) signal from noisy single channel noisy electromyography (EMG) signal has been developed. The method is based on singular spectrum analysis, where the subspace of interest is estimated by incorporation of a priori information. The method is applied to a number of real data sequences. The performance of the algorithm is both subjectively and objectively assessed and the results are compared with those of the well known adaptive line enhancer (ALE) since at least one of the components, i.e. ECG, is quasi-periodic. It is demonstrated that the proposed subspace method significantly outperforms ALE.