An Accurate, Versatile, and Robust Brain Switch for Neurorehabilitation

Abstract

AbstractFor the past decade, our group worked towards the development of a non-invasive BCI system for neuromodulation. Until recently, BCIs have been used mainly for communication and replacement or restoration of lost functions for severely disabled people. Using a BCI for neuromodulation requires that the protocol closely matches the steps involved in the motor learning process. However, the underlying mechanisms of motor learning in humans remain elusive, though several possibilities have been proposed. Of these, the most promising was proposed by Hebb (The organization of behavior: a neuropsychological theory, vol. 44, p. 335, 1949), who suggested that synaptic strength is increased when two inputs from two sources arrive at the post-synaptic cell in synchrony. If this occurs repetitively with the necessary intensity, synaptic strength is increased. That is, the same input will produce a greater output. Stefan et al. (Brain J Neurol 123(pt 3):572–584, 2000) were the first to investigate this concept non-invasively in humans, and it has now become accepted that it closely matches what occurs during motor learning. With this knowledge, we developed a BCI system where the user’s movement intention is detected through non-invasive electroencephalogram (EEG). When the onset of the intended movement is detected, it is used to drive an external device that produces the intended movement. Through this process, the user is provided with the necessary proprioceptive feedback, timed to coincide with the onset of the intended movement, so that the Hebbian principle of associativity is satisfied. In this chapter, we outline the development of this BCI system for neuromodulation and show that it can be used to drive any external device while satisfying all the main criteria necessary for a full BCI application, namely: accuracy, flexibility, rapid control, and robustness.KeywordsBrain computer interfaceHebbian plasticityMovement related cortical potentialsNeurorehabilitation