Distributed retinal stimulation model based on adaptive system

We propose in this paper a new methodology to generate the output signals of an artificial retinal stimulator to selectively stimulate the human optic nerve. The output signal generator is based on clinical electrical patterns recorded from a multi-focal electroretinography. An adaptive filter is proposed to estimate the weights of a set of finite impulse response (FIR) filters, that generate each electrical response highly correlated with the visual patterns. The proposed model let to analyse the behaviour of the retina coded signals with a set of linear FIR filters.