Home // ACHI 2019, The Twelfth International Conference on Advances in Computer-Human Interactions // View article

Assessing the Impact of Muscular Fatigue on Myoelectric Signals Using Myo Armband

Authors:
Sudhir Sharma
Volker Steuber
Farshid Amirabdollahian

Keywords: Gesture detection; fatigue; electromyography.

Abstract:
This paper investigates the impact of muscular fatigue on myoelectric signals in order to incorporate biological feedback acquired wirelessly into a game interaction intended for rehabilitation. The study was conducted in four phases: Familiarisation, Training-1, Dumbbell exercises and Training-2 with 20 healthy participants. During the two training phases, each participant performed 5 hand gestures, each gesture was performed in 5 iterations. Each iteration lasted 5 seconds and consisted of 5 repetitions of each gesture, visually guided to last a second each. The logged data was recorded into CSV (Comma-Separated Values) files at a rate of 200Hz. In order to compare grasps before and after the fatiguing exercise, we used SVM (Support Vector Machine), using the capability to judge grasp accuracy for each phase. By comparing the grasp accuracy pre and post exercise, we found that muscular fatigue can negatively impact on gesture accuracy. The gesture accuracy detected after dumbbell exercises was significantly lower than that of the gestures performed before exercise. The data collected through subjective questionnaires also supports the results. In this study, we identified that fatigue, caused by interaction intensity and effort, can affect the accuracy of gesture detection. This has a further implication for cases where EMG (Electromyography) is used as the biofeedback involved in human-machine interactions, such as gameplay. While it is thought that more intensive interaction may benefit recovery after stroke, it is possible to optimise interactions towards reducing fatigue, for example by pacing the game difficulty based on detected level of fatigue.

Pages: 159 to 164

Copyright: Copyright (c) IARIA, 2019

Publication date: February 24, 2019

Published in: conference

ISSN: 2308-4138

ISBN: 978-1-61208-686-6

Location: Athens, Greece

Dates: from February 24, 2019 to February 28, 2019