The major reason for the high rate of rejection of myoelectric prostheses after upper arm amputation is the unnatural control of the hand and wrist functions as well as the limited number of degrees of freedom (DoF) of the prosthesis. No solution has ever been proposed based on the natural neuromuscular reorganization after amputation leading to a mobile phantom limb in 77% of upper limb amputees. It had been shown that phantom movements are associated to specific activities of residual limb muscles according to the type of performed phantom movement. Interestingly, we found that neuromuscular reorganization gives rise to the genesis of functionally distinct muscle volumes within the same muscle, i.e., muscle volumes that are distinctly and independently active as function of the type of phantom movement. If we use these patterns of activation of the distinct muscle volumes to make the prosthesis mimicking the executed phantom movement, the control would be natural for the patients and the number of controllable active DoF of the prosthesis might potentially increase. So, the main goal of PhantoMovControl is to outreach the actual limits of myoelectric upper-arm prostheses by using high-resolution EMG associated with phantom movements, favouring a natural and intuitive control of prosthetic devices (thus without heavy training/learning) with a high number of DoF. Our main ambition is to develop within 3 years a “control kit” for existing polydigital hand prostheses and to rapidly transfer this approach to both clinic and to prosthesis industry. This control kit must integrate arrays of electrodes that can be placed on the residual limb within the socket of the prosthesis (thus ultra-thin, cable free, and fixed on an extensible support), as well as an embedded computer module that classifies muscle activity associated with phantom movements and controls in real time the polydigital prosthesis. Multi-electrode arrays are used in the literature and some are commercially available, but they do not fulfil our requirements. Therefore, we integrated an industrial partner able to design and produce such electrodes arrays to our academic consortium in order to accelerate the transfer of our new control approach to a concrete implementation by prosthetics companies in their products. In order to make our bio-inspired control approach robust and adaptable to intra- and inter-patient variability, we’ll have to progress in parallel at a fundamental level. We’ll study the neuromuscular reorganization after amputation and gather knowledge about the robustness of the EMG during different types of phantom movements (e.g., combined with residual limb movements, fast or slow…) and their evolution in time and following phantom movement training. Moreover, we’ll study the socio-anthropological and cultural phenomena influencing the appropriation of these technical objects in order to ease the appropriation by the patients of such myoelectric prostheses and reduce the rate of rejection of upper-limb prostheses. To ensure the multidisciplinarity of our approach, partners from Human and Social Sciences, Neurosciences and Engineering were gathered into the project. The consortium consists of 3 scientific (ISM, ISIR, CETCOPRA), 1 clinical (IRR) and 1 industrial partner (Microvitae), closely interacting in 5 Work Packages. Requested resources include an engineer to assist in the development of the EMG classifier and the “control kit” (ISIR), a postdoctoral researcher for studying the robustness of the EMG and its evolution after training (ISM), a prosthesist and an occupational therapist for adaptation of the prosthesis and its functional evaluation (IRR). We take up the challenge that this new control approach will function symbiotically with the underlying neuroplasticity, hereby stabilizing the control of the phantom movements and the associated EMG (and thus of the prosthesis) and increasing the appropriation by the patients of such technology.
Madame Jozina De Graaf (Institut des Sciences du Mouvement)
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
INST REG DE READAPTATION DE NANCY
Université Paris 1 Centre d'Etudes des Techniques des Connaissances et des Pratiques
ISM Institut des Sciences du Mouvement
ISIR-UPMC Insititut des Systèmes Intelligents et de Robotique
Help of the ANR 507,176 euros
Beginning and duration of the scientific project: September 2015 - 36 Months