Exploration of motor unit activation pattern using high density sEMG minimum norm estimation

when speaking of the electrical activity of the muscles, major data can be extracted from such signals. One significant piece of information that can be extracted from those signals is the spatial positioning of the active regions within the muscle with major applications it participate in, starting from the improved diagnosis of neuromuscular diseases, and ending in a better prosthetic control. This study is an extended and improved version of a recent work previously done that included proposing the creation of a modified lead field matrix for High-Density surface EMG signals. This was proceeded by solving the corresponding inverse problem by the classical methods of minimum norm estimation (regular and weighted) and the comparison of the two methods by considering two different scenarios of activation patterns; one which only takes into consideration the superficial sources and the other that takes into consideration the deep sources. Then, the mean square error between the signals reconstructed using the inverse solution and the initial signals, which are simulated by an analytical model, is calculated for both of the scenarios.