Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Without Surgical Reinnervation: A Preliminary Study

• Main Authors: Nathanaël Jarrassé, Etienne de Montalivet, Florian Richer, Caroline Nicol, Amélie Touillet, Noël Martinet, Jean Paysant, Jozina B. de Graaf
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2018-11-01
• Series: Frontiers in Bioengineering and Biotechnology
• Subjects: prosthetics; transhumeral amputation; phantom limb; myoelectric control; pattern recognition; voluntary phantom limb mobility
• Online Access: https://www.frontiersin.org/article/10.3389/fbioe.2018.00164/full

Transhumeral amputees face substantial difficulties in efficiently controlling their prosthetic limb, leading to a high rate of rejection of these devices. Actual myoelectric control approaches make their use slow, sequential and unnatural, especially for these patients with a high level of amputation who need a prosthesis with numerous active degrees of freedom (powered elbow, wrist, and hand). While surgical muscle-reinnervation is becoming a generic solution for amputees to increase their control capabilities over a prosthesis, research is still being conducted on the possibility of using the surface myoelectric patterns specifically associated to voluntary Phantom Limb Mobilization (PLM), appearing naturally in most upper-limb amputees without requiring specific surgery. The objective of this study was to evaluate the possibility for transhumeral amputees to use a PLM-based control approach to perform more realistic functional grasping tasks. Two transhumeral amputated participants were asked to repetitively grasp one out of three different objects with an unworn eight-active-DoF prosthetic arm and release it in a dedicated drawer. The prosthesis control was based on phantom limb mobilization and myoelectric pattern recognition techniques, using only two repetitions of each PLM to train the classification architecture. The results show that the task could be successfully achieved with rather optimal strategies and joint trajectories, even if the completion time was increased in comparison with the performances obtained by a control group using a simple GUI control, and the control strategies required numerous corrections. While numerous limitations related to robustness of pattern recognition techniques and to the perturbations generated by actual wearing of the prosthesis remain to be solved, these preliminary results encourage further exploration and deeper understanding of the phenomenon of natural residual myoelectric activity related to PLM, since it could possibly be a viable option in some transhumeral amputees to extend their control abilities of functional upper limb prosthetics with multiple active joints without undergoing muscular reinnervation surgery.