Functional connectivity in primary visual cortex

To help elucidate how orientation tuning is created in V1, one may examine whether sharper tuning is coincident with certain regions of the V1 architecture. For instance, V1 orientation maps provide a continuum of functional contexts that could account for differences in the selectivity of individual cells. Some cells lie in regions of more homogeneous orientation preference (iso-domains) while others lie in regions with a variety of preferences (e.g. pinwheel centers). This was the motivation behind our first experimental design where we used a novel combination of multielectrode arrays and optical imaging to precisely determine the location of electrodes within orientation maps. The method revealed that the most selective cells for orientation tend to lie in the most homogeneous regions of the orientation map. By pooling data from both cats and monkeys, we also found that there is a consistent relationship across the two species. This suggests that sharper tuning of cells in the cat than in the monkey can be accounted for by differences in the overall structure of their orientation maps. The identification of regions within the functional architecture that generate the most sharply tuned cells provides further constraints on the precise mechanisms by which tuning arises.;Our next question entails the relative contributions from the thalamus and the extensive V1 lateral network in producing V1 responses. Using 10x10 electrode arrays implanted in V1 we simultaneously recorded spikes and local field potentials (LFP) at all electrodes. Computing the spike-triggered LFPs revealed that spikes generate a traveling wave of synaptic input whose rate is consistent with previous estimates of conduction velocity in cortical tissue. The spatial amplitude decay of this traveling wave was used to assess the extent of lateral connectivity. When the stimulus is absent or weak, lateral input over a large area of cortex strongly influences local population activity. As the strength of the visual stimulus increases, lateral input is reduced in overall magnitude and spatial extent. These findings provide a new conceptual framework of cortical computation that reconciles opposing positions on the relative impact of the lateral network on visual responses.