Force Myography （FMG）, which monitors pressure or radial deformation of a limb, has recently been proposed as a po- tential alternative for naturally controlling bionic robotic prostheses. This paper presents an exp...Force Myography （FMG）, which monitors pressure or radial deformation of a limb, has recently been proposed as a po- tential alternative for naturally controlling bionic robotic prostheses. This paper presents an exploratory case study aimed at evaluating how FMG behaves when a person with amputation uses a hand prosthetic prototype. One volunteer （transradial amputation） participated in this study, which investigated two experimental cases： static and dynamic. The static case considered forearm muscle contractions in a fixed elbow and shoulder positions whereas the dynamic case included movements of the elbow and shoulder. When considering eleven different hand grips, static data showed an accuracy over 99%, and dynamic data over 86% （within-trial analysis）. The across-trial analysis, that takes into account multiple trials in the same data collection set, showed a meaningful accuracy respectively of 81% and 75% only for the reduced six grips setup. While further research is needed to increase these accuracies, the obtained results provided initial evidence that this technology could represent an in- teresting alternative that is worth exploring for controlling prosthesis.展开更多
文摘Force Myography （FMG）, which monitors pressure or radial deformation of a limb, has recently been proposed as a po- tential alternative for naturally controlling bionic robotic prostheses. This paper presents an exploratory case study aimed at evaluating how FMG behaves when a person with amputation uses a hand prosthetic prototype. One volunteer （transradial amputation） participated in this study, which investigated two experimental cases： static and dynamic. The static case considered forearm muscle contractions in a fixed elbow and shoulder positions whereas the dynamic case included movements of the elbow and shoulder. When considering eleven different hand grips, static data showed an accuracy over 99%, and dynamic data over 86% （within-trial analysis）. The across-trial analysis, that takes into account multiple trials in the same data collection set, showed a meaningful accuracy respectively of 81% and 75% only for the reduced six grips setup. While further research is needed to increase these accuracies, the obtained results provided initial evidence that this technology could represent an in- teresting alternative that is worth exploring for controlling prosthesis.
