Thalamic dis-inhibition strengthens the retinal inhibitory surrounds at the lateral geniculate nucleus

A computational theoretic approach is adopted to show that the thalamic dis-inhibitory pathway of an alert cat strengthen the retinal ganglion cells' inhibitory surrounds at the dorsal lateral geniculate nucleus. First, a criterion is introduced to evaluate the sharpness of a two-dimensional activity pattern of cells residing at a rectangular lattice. Then, a parallel iterative algorithm is introduced as the computational mechanism of the dLGN circuitry. Probed by an activity pattern, the algorithm evolves in order to maximize the sharpness criterion. Finally, a modular neural network compatible with the known circuitry of the cat's thalamic dis-inhibitory pathway is introduced to implement the algorithm.