Outline of a theory for the ontogenesis of iso-orientation domains in visual cortex

The following statements are assumed as experimental facts: The majority of cells in visual cortex is selectively sensitive to stimuli in the form of lines or edges of a certain orientation. Optimal orientation varies smoothly from cell to cell over the cortical surface. This structure is present, in immature form, at the time of eye opening. The problem is to find a plausible mechanism for the ontogenesis of this structure. Published theories are reviewed and found to be insufficient in one or more respects. A new theory is proposed. It distinguishes two processes: A) In the previsual period, intracortical connections re-organize themselves such that they restrict activity to an appropriate family of patterns (e.g. stripes of different phases). B) Later, a one-to-one mapping is developed that couples each stimulus orientation to one of the cortical patterns. The coupling may take place in two stages: a) Still before eye opening, two sets of afferent pilot fibres serve to fix the mapping in two points. (The pilot fibres can be assigned to these two sets on the basis of correlated or anticorrelated spike activity. They have genetically programmed orientation preferences which fall into two groups. A fibre sorting mechanism analogous to the one producing ocularity domains concentrates the sets of fibres into complementary activity patterns.) b) After eye opening, a more precise mapping of orientations to cortical patterns develops with the help of visual stimulation and synaptic plasticity according to a well-known mechanism. Experimental evidence for the theory is discussed. The theory can be evaluated in terms of general principles of brain organization.Supported in part by NATO grant No. 1791