CNN models of receptive field dynamics of the central visual system neurons

Deals with the biological aspects of the receptive field (RF) concept and its possible cellular neural network (CNN) modeling. Three kinds of receptive field definitions are discussed: the experimentally measured RF, the mathematical model of the RF and its anatomical background. Previously, new RF-mapping techniques have revealed that neurons in the visual pathway exhibit striking RF dynamics, which implies that for adequate characterization the RF profile has to be examined in the space-time domain. Starting from these findings in the present study the neurons' static RF definition is purified and some experimental results of De Angelis et al. (1995) are modeled by the CNN. Our CNN model indicates that the spatio-temporal RF dynamics can be generated by time invariant synaptic strength values.