Motor unit action potentials as a source of noise in the non-invasive detection of fibrillation potentials

Denervated muscle fibers produce spontaneous depolarizations termed fibrillation potentials. These potentials are an indicator of neuromuscular pathology and are detected by inserting a needle electrode into the muscle of interest to detect the time-based signal. A proposed noninvasive method measures the spectral energy corresponding to increased spontaneous muscle activity. This paper examines the impact of normal muscle activity on such a method through the use of a computer model of fibrillation potentials and normal motor unit action potentials. A mathematical expression for the surface recorded signal is proposed and used as the basis for analyzing the temporal and spectral characteristics of spontaneous and normal motor activity. Based on these results, filtering methods for the removal of normal motor activity are proposed and future work needed to implement non-invasive detection of fibrillation potentials is discussed.