Identification of motor unit spatial activation by minimum norm estimation

In the context of exploring the internal muscle activity, information such as the location of the muscle electrical sources play a vital role in the diagnosis of neuromuscular diseases, biomechanics, medical robotics research, and better prosthetic control. In this paper the inverse problem of surface electromyography (sEMG) source estimation, tested on simulated data only, was solved by two approaches; the regularized minimum norm estimation (MNE) and the weighted minimum norm (wMNE) estimation. Using a single layer volume conductor, we compared simulated sEMG signals to reconstructed signals after the application of the inverse problem by the two methods. The novelty in the work is that those two methods are compared on a sEMG adapted lead field. Results showed the effectiveness of the MNE and wMNE approaches according to mean square error (MSE) when comparing the original signal and the reconstructed signal.