Removal of Electrocardiogram Interference from Diaphragmatic Electromyogram Signals using Sliding Singular Spectrum...

Over recent years, Singular Spectrum Analysis (SSA) has gained popularity as an effective means to denoise biologically sourced single channel signals, especially Electromyogram (EMG) and Electrocardiogram (ECG) signals amongst others. There are numerous applications whereby the signal acquisition process results in the mixing of both types of signals along with body motion artifacts and the inevitable electromagnetic interference. Both ECG and EMG signals are very useful to physicians, though preferably in isolation, though they rarely present themselves in this manner. Simple filtering techniques are ineffective in their separation as both signal spectra overlap in the frequency domain. In this paper, we propose a technique based on a sliding SSA algorithm which proves to be more successful in separating real mixed EMG and ECG signals than traditional block based approaches on single channel data. SSA is a non-parametric technique that decomposes the original time series into a number of additive components, each of which can then be readily identified based on statistical analysis as belonging to EMG or ECG signals. This approach could be applied equally to other signal types using different statistical methods as required, moreover, this technique is relatively straight-forward to implement and does not require any reference signals or training.