Myoelectric activity detection during a Sit-to-Stand movement using threshold methods

  • Authors:

  • Ghulam Rasool;Kamran Iqbal;Gannon A. White

  • Affiliations:

  • Systems Engineering Department, University of Arkansas at Little Rock, 72204 Little Rock, AR, USA;Systems Engineering Department, University of Arkansas at Little Rock, 72204 Little Rock, AR, USA;Health Sciences Department, University of Arkansas at Little Rock, 72204 Little Rock, AR, USA

  • Venue:
  • Computers & Mathematics with Applications
  • Year:
  • 2012

Quantified Score

Hi-index 0.09

Visualization

Abstract

Myoelectric signals recorded via surface electrodes contain rich muscle activity information that is beneficial for both clinical diagnosis and biomedical research. When synchronized with the kinematic data, these signals provide investigators an insight into muscle activation sequence, onset, levels, and periods. A primary difficulty with the analysis and interpretation of electromyography (EMG) signals lies in the inherent stochastic nature of the EMG process, which arises from its biological variability as well as noise added during the collection process. Various techniques for muscle onset and activity detection from the myoelectric signal have been proposed in the literature. Our focus in this study is myoelectric activity detection from EMG signals collected during Sit-to-Stand (STS) and Stand-to-Sit (STST) movements. We explore a previously established double threshold detection method, and compare its results with a novel detection scheme based on the energy of the signal. Accordingly, EMG signals from four lower extremity muscles, and synchronized kinematic data, were collected for 180 trials of STS and STST movements performed in the laboratory. Detection thresholds above baseline in the case of both algorithms were computed and analyzed using a 2 (detectors) x 4 (activity thresholds) repeated measures analysis of variance. Our statistical analysis revealed that the energy detection method performed similarly to the double threshold method, while both methods required a considerably higher threshold above baseline for detection.