Berlin Lectures on Neurotechnology: Towards High Performance, Weakly Invasive Brain Computer Interfaces Using Selective Visual Attention

Brain-Computer-Interfaces have been proposed as a solution for paralyzed persons to communicate and interact with their environment. However, the neural signals used for controlling such prostheses are often noisy and unreliable, resulting in a low performance of real-world applications.

Here we propose neural signatures of selective visual attention as a fast, reliable, and high-performance control signal for brain prostheses. We recorded epidural field potentials with chronically implanted electrode arrays from visual cortex of two macaque monkeys engaged in a shape-tracking task, and classified the direction of attention to one of the two visual stimuli. Classification performances ranged up to 99.9% and information transfer rates achieved up to 242 bits/min. The most informative signal feature was spectral amplitude in the gamma band but also phase coherence was highly informative. Since our epidural recordings were stable over periods of more than 36 months our findings suggest a novel paradigm for constructing robust brain prostheses. For this purpose we are now developing wireless high density arrays that would allow lifetime implantation in patients suffering severe handicaps.