Multi-postural myoelectric control of 3D-printed transradial prosthesis

Abstract

Improvement in the control strategy for transradial prosthetics is among the top design priorities for prosthesis users [1]. Conventional control schemes for prosthetic hands only provide limited number of movements; usually basic opening and closing for rudimentary clamping [2]. To address this gap, this study presents two schemes (M1-A and M1-B) as myoelectric control strategies to facilitate multi-grasp/postural control of a transradial prosthesis. The proposed control method supports traversing through a set of preprogrammed postures, contextual-locking of fingers, and grasp-specific finger flexion/extension adjustment. Additionally, a data glove control interface (M2) was also developed and was used to characterize the performance of the native hand and set a baseline for comparison for the proposed myoelectric control schemes. The prototype was evaluated in terms of aspects of controllability. The results of the quasi-experiments showed that the developed multigrasp myoelectric control strategies have an average posture transition time and a grasp transition success rate of 2.306s and 96.4% for M1-A; while M1-B achieved 1.727s and 99.03%, respectively. Both schemes achieved an above average rating in perspicuity based on the UEQ. Relative to the characterized native hand control performance, the overall test results for the proposed myoelectric control schemes are both deemed as satisfactory, with M1-B performing slightly better.