Understanding the importance of natural neuromotor strategy in upper extremity neuroprosthetic control.

Abstract

A key challenge in upper extremity neuroprosthetics is variable levels of skill and inconsistent functional recovery. We examine the feasibility and benefits of using natural neuromotor strategies through the design and development of a proof-of-concept model for a feed-forward upper extremity neuroprosthetic controller. Developed using Artificial Neural Networks, the model is able to extract and classify neural correlates of movement intention from multiple brain regions that correspond to functional movements. This is unique compared to contemporary controllers that record from limited physiological sources or require learning of new strategies. Functional MRI (fMRI) data from healthy subjects (N = 13) were used to develop the model, and a separate group (N = 4) of subjects were used for validation. Results indicate that the model is able to accurately (81%) predict hand movement strictly from the neural correlates of movement intention. Information from this study is applicable to the development of upper extremity technology aided interventions.