Classification of low-level finger contraction from single channel surface EMG

This paper reports a study that has investigated a new technique to identify very low level finger flexion by the classification of single channel surface electromyogram (sEMG). The technique is based on decomposing of sEMG based on the model of transmission of motor unit action potentials (MUAP) in body tissues. This technique is relevant for identifying control commands that are often based on low level and complex muscle contraction such as finger flexion which are often a convenient way for a user to control equipment or a prosthesis device. Use of single channel is extremely important because it does not require an expert to mount the electrodes and has a further advantage in reduced cost and computational complexity. The paper reports experiments conducted on four healthy volunteer subjects with four actions and tested over 50 repetition and a high classification accuracy.