Visual cortex prostheses and neurotechnology

Communicating with the brain via a neuroprosthesis is a challenging endeavour. On the technological side, it is difficult to establish reliable and long-lasting interfaces to neurons and neuron populations, and it is even more demanding to selectively activate specific target cells or cell assemblies. But even if these technical issues could be solved, there is a computational problem which is often neglected, but of equal importance:

Mosaic Patterns
Understanding what’s going on in cortex: Modelling transient, patterned activation states in the visual system.

Ongoing activtiy and the language of the brain

With increasing distance from the sensory periphery, neural representations are more and more shaped by the ongoing computations in brain networks. For knowing where and when to stimulate, or for interpreting neural activities correctly, one has to understand the ‘language’ of the brain. For improving neuroprostheses, we believe it to be essential to better understand that language, in particular if one is targeting brain areas in visual cortex. Furthermore, it is important to listen what is already going on in brain networks, and to tailor neurostimulation to that on-going activity rather than forcefully trying to ‘overwrite’ these activation states.

ISee Scheme
Improving communication with the brain: The I-See framework for bi-directional visual cortex prostheses.

I-See: Improving visual cortex prostheses for the blind

For bringing these theoretical ideas to fruition, we are collaborating with an international group of scientists to develop better cortical visual prostheses for the blind. For further information, see the “I-See” project homepage.