Universal pinwheel statistics in the visual cortex

The organization of orientation columns into pinwheel-like patterns has been observed in a wide variety of animal species including galago, ferret, and tree shrew. These mammals have been separated for more than 50 million years of evolution, occupy different ecological niches, and exhibiting distinct patterns of visual behavior. Consequently, many features of their visual systems differ substantially. Here we show that despite this divergence, basic statistics of the pinwheel pattern are universal in the primary visual cortex of galagos, ferrets, and tree shrews. We analyzed the spatial organization of pinwheels in orientation maps obtained by intrinsic optical imaging using a novel pinwheel analysis method that is robust against noise. In particular, we focused on the pinwheel density, i.e. the mean number of pinwheels per area with linear extent of one column spacing. In 26 tree shrew hemispheres, the average pinwheel density was 3.12 (0.04) [mean (s.e.m.)] with mean pinwheel densities in individual maps ranging from 2.7 to 3.5. In 9 galago hemispheres, pinwheel densities varied comparably with an average of 3.18 (0.09). An average pinwheel density of 3.16 (0.03) was found in a sample of 82 ferret hemispheres with values ranging between 2.0 and 4.0 for individual ferrets. Thus, the average pinwheel density was indistinguishable in the three species. The total average was 3.14 (0.03). The variation among different hemispheres was mainly determined by the typical size of the analyzed regions in the maps. In addition to the density, we observed almost identical nearest neighbor statistics of pinwheels in all three species. Theoretical analyses show that the observed universal statistics are quantitatively reproduced by models of cortical self-organization dominated by long-range interactions. We conclude that the experimentally observed universal pinwheel statistics are emergent signatures of a dominant role of long-range interactions in the self-organization of cortical circuits.