Novel Methods in SEMG-Force Estimation

Human body motion relies heavily on muscle contractions. Reliable measurement of the intensity of muscle contraction is required in a number of applications including control of prostheses, sports medicine, human-robot interaction and medical rehabilitationThe surface electromyogram (SEMG) is a non-invasive indirect technique for muscle force estimation. However, even for simple isometric contractions, SEMG is affected by physiological and non-physiological factors, which impact the accuracy of force estimation using this technique. Accurate force estimation is even more complicated for dynamic contractions as additional factors such as contraction velocity and varying motor unit activation patterns during the contraction impact force generation. To accurately estimate muscle forces using SEMG signals two fundamental steps are necessary: first, precise EMG amplitude estimation, and second a modelling scheme capable of coping with the non-linearities and dynamics of the EMG-force relationship. In this book, angle-based EMG amplitude calibration and parallel cascade identification (PCI) modelling have been combined to address the issue of EMG force estimation in dynamic contractions.