Improving Wrist Force Estimation With Surface EMG During Isometric Contractions

Abstract

In this paper, methods for selecting channels to improve estimated force using fast orthogonal search (FOS) have been investigated and a new method is proposed. The surface electromyogram (sEMG) signal acquired from linear surface electrode arrays, placed on the long head and short head of biceps brachii and brachioradialis during isometric contractions are used to estimate force induced at wrist using the FOS algorithm. In this paper, the effects of the sEMG signal characteristics obtained from the arrays and channels’ locations on the estimated force are investigated to find channels resulting in force estimation improvement compared to using all available channels. Several methods for channel selection have been studied, showing that the sensitivity of the estimated force to the location of the channels is subject-dependent. The proposed method uses only the channels with highest mean of power spectrum density (PSD) and low cross correlations. The channels selected by this method have improved FOS force estimate compared to using all the available channels.