Determining the Time until Muscle Fatigue using Temporally Extended Prediction Learning
Full Text: Pilarski_2013_IFESS_Postprint.pdfAnticipating the time remaining until muscle fatigue during the real-time use of an assistive rehabilitation device promises to improve the health and mobility of patients. In this work, we present a data-driven machine learning approach for assessing and reporting on situation-specific muscle fatigue in a point-of-care setting. The time until the onset of fatigue is predicted from interactions between a user and their wheelchair, as captured through myoelectric signals and pushrim data from a SmartWheel recording system. Our findings indicate that a real-time learning approach is able to accurately forecast the time remaining until a subject reaches fatigue-related activity thresholds. Our results also suggest the potential for generalizing these personalized anticipations between different patients. The approach presented in this work therefore promises to allow both a user and their device’s control system to observe endurancerelated future effects of current motor control choices; ongoing feedback of this kind may prove to be a valuable tool for improving patient mobility outside of the clinical environment.
Citation
P. Pilarski, L. Qi, M. Ferguson-Pell, S. Grange. "Determining the Time until Muscle Fatigue using Temporally Extended Prediction Learning". International Functional Electrical Stimulation Society Conference, pp 37–40, June 2013.Keywords: | |
Category: | In Conference |
Web Links: | UofA |
BibTeX
@incollection{Pilarski+al:IFESS13, author = {Patrick M. Pilarski and Liping Qi and Martin Ferguson-Pell and Simon Grange}, title = {Determining the Time until Muscle Fatigue using Temporally Extended Prediction Learning}, Pages = {37–40}, booktitle = { International Functional Electrical Stimulation Society Conference}, year = 2013, }Last Updated: January 19, 2021
Submitted by Sabina P