Optimal electrode configurations for EMG experiments using source separation techniques

Surface Electromyogram (sEMG) is a ubiquitous tool which can be used to estimate the muscle action potential propagation velocity. Its analysis permits a better knowledge of the muscle activation sequence either in the lower or upper extremities, during predetermined voluntary movements. In the most general form, sEMG electrodes are used to evaluate muscle activity for function, control and learning. Multiple electrodes are essential for extracting intrinsic physiological and contextual information in using sEMG signals. Due to the heterogeneous nature of the Electrodes compounded by the mixture of signals across different electrode channels, it would be practically difficult to obtain useful information. Hence there is need for method to indicate the quality of the electrode set up during the sEMG based experiments. The purpose of this paper is to describe the use of source separation for sEMG based on Independent component analysis (ICA). We demonstrate how this can be used in practical sEMG experiments when the number of sensing channels is varied.