Locomotion mode classification using force myography

In this paper, we developed a force myography (FMG) system to capture the volumetric changes in thigh muscles during locomotion and applying it to recognize the current locomotion mode. The force myographic signal is acquired from eight sensors wrapped around the thigh. Six healthy volunteers participated in the study and five locomotion modes i.e. level walk (LW), ramp ascent (RA), ramp descent (RD), stair ascent (SA) and stair descent (SD) were tested in the study. To evaluate the classification performance of the FMG system, linear discriminant analysis (LDA) classifier was used. The average classification accuracy for locomotion modes was 92.5±7.1% with 97.2% classification accuracy for stair ascent walking.