Autocorrelation analysis of lower limb EMG signals for the initial evaluation of hemiparetic gaits

Abstract

In order to investigate evaluation methodology for hemiparetic gaits on rehabilitation purpose, volunteering subjects are recruited on the standardized walking trials. EMG signals are collected from six muscle groups on each lower limb. Improved autocorrelation analysis combined with gait analysis is then proposed. As the completion of gait cycle is described by percentage, gait quality is evaluated by the periodic characteristics and the peak value occurs on 100 %, 200%, 300%, and 400% of gait cycles from zero phase that are recorded. The similarities are assessed by specific number from 0 to 1.0, and the more approach to 1.0, the higher similarity is represented. Then the function of balance control ability is assessed between left and right sides. The resulting evaluation of gait qualities are initially assessed based on measuring and numerical outcome. The effect of the proposed method is verified by a case study as it is clearly agreed with the qualitative evaluation obtained from clinical professionals. More clinical studies will be carried out to further confirm the effectiveness of this proposed method. The proposed evaluation methodology could also be potentially used as the bio-feedback for the robotic-assisted rehabilitation in future.