Extending Myoelectric Prosthesis Control with Shapable Automation: A First Assessment

For many users of myoelectric prostheses there is a set of functionality which remains out of reach with current technology. In this work, we provide a first assessment of an extension to classical myoelectric prostheses control approaches that introduces simple automation that is shapable, using EMG signals. The idea is not to replace classical techniques, but to introduce automation for tasks, like those which require the coordination of multiple degrees of freedom, for which automation is well-suited. A prototype system is developed in simulation and an exploratory user study is performed to provide a first assessment, that evaluates our proposed approach and provides guidance for future development. A comparison is made between different formulations for the shaping controls, as well as to a classical control paradigm. Results from the user study are promising: showing significant performance improvements when using the automated controllers, and also unanimous preference for the use of automated controllers on this task. Additionally, some questions about the optimal user interaction with the system are revealed. All of these results support the case for continued development of the proposed approach, including more extensive user studies.