Amplitude indicators and spatial aliasing in high density surface electromyography recordings

Abstract

Abstract Average Rectified Value (ARV) and Root Mean Square (RMS) are amplitude indicators commonly used in the field of EMG either in time or space. These two indicators are compared (a) analytically for a one dimensional sinusoid, sum of sinusoids, two dimensional sinusoids, and (b) numerically by simulating a high density detection system, sampling in space the distribution of propagating surface action potentials generated by a muscle motor unit (MU). For any signal sampled above the Nyquist frequency the estimated RMS does not depend on the sampling rate while the estimated ARV does. The surface potential is often sampled in space below the Nyquist frequency, by high density surface EMG detection systems (HDsEMG), generating aliasing in space. For point-like electrodes, the lowest spatial sampling frequency corresponding to the largest inter-electrode distance (IED), which avoids spatial aliasing for a simulated MU action potential, is 100 samples/m (IED = 10 mm). Therefore, IEDs below this value are recommended for measurements of EMG image features. From the theoretical point of view, the spatial RMS of sEMG images is more robust than the ARV with respect to the IED and should be preferred.