Translating Image Sequences into Spike Patterns for Cortical Neuro-Stimulation

This paper describes a bio-inspired preprocessing and coding system devised for producing optimal multielectrode stimulation at the cortical level, starting from image sequences, and working at video rates.The system is implemented as a hybrid platform with software and reconfigurable hardware. It produces as output a continuously varying stream of pulses or spike patterns. The main objective of this work is to build, in a near future, a portable system for a visual neuro-prosthesis able to stimulate efficiently an array of intra-cortical implanted microelectrodes. Thus, a set of parameters can be adjusted in the designed processing and spike coding modules in order to trade-off their technology constraints with the biological plausibility of their functional features. Spike timing and possible temporal correlation among different microelectrodes are preserved as important cues to improve the stimulation efficiency and/or to initiate plasticity processes at the receiver neurons in the primary visual cortex