The network behind spatio-temporal patterns: building low-complexity retinal models in CNN based on morphology, pharmacology and physiology

In this paper, a vertebrate retina model is described based on a cellular n eural network (CNN) architecture. Though largely built on the experience of previous studies, the CNN computational framework is considerably simplifi ed: first-order RC cells are used with space-invariant nearest-neighbour in teractions only. All non-linear synaptic connections are monotonic continuo us functions of the pre-synaptic voltage. Time delays in the interactions a re continuous represented by additional first-order cells. The modelling ap proach is neuromorphic in its spirit relying on both morphological and phar macological information. However, the primary motivation lies in fitting th e spatio-temporal output of the model to the data recorded from biological cells (tiger salamander). In order to meet a low-complexity (VLSI) implemen tation framework some structural simplifications have been made. Large-neig hbourhood interaction (neurons with large processes), furthermore inter-lay er signal propagation are modelled through diffusion and wave phenomena. Th is work presents novel CNN models for the outer and some partial models for the inner (light adapted) retina. It describes an approach that focuses on efficient parameter tuning and also makes it possible to discuss adaptatio n, sensitivity and robustness issues on retinal 'image processing' from an engineering point of view. Copyright (C) 2001 John Wiley & Sons, Ltd.