Patterns appear on visual cortex as a result of hallucinations, migraines, other disruptions of normal processing of visual input. Such patterns may provide a clue to spatial neural structure and dynamics of the cortex. This talk reports on an effort to model visual cortex structure in terms of reaction-diffusion equations with discrete Laplacian, which are known to produce Turing patterns and, in the presence of stochasticity, produce quasi-patterns, meaning otherwise damped patterns sustained by noise.