Abstract
The slightly different signals coming from the two eyes allow our brain to compute a tridimensional representation of the visual world. The monocular inputs reach the cortex in layer 4 where they are segregated and form ocular dominance domains. Then they are first combined by binocular neurons in layers 2/3. At the surface of the cortex, ocular dominance domains are spatially organized in so-called ocular dominance stripes. Studying how the information coming from the two eyes is integrated at the mesoscopic scale is of relevance. For this purpose, we used optical imaging in anesthetized monkey to analyze how these signals are integrated and summed at the population level in V1. Ocular dominance maps were obtained with standard intrinsic optical imaging procedures. Dichoptic interactions were studied with voltage-sensitive dye imaging (VSDI) using a frequency-tagging paradigm. Visual stimuli of different contrasts were presented at 6Hz and 10Hz to the left and right eye respectively, separately or simultaneously with a passive 3D monitor. The frequency analysis of the evoked response was used to identify the contribution of each eye. We observed that the population activity in V1, elicited by the stimulation of one eye, is suppressed by a dichoptic stimulation. This signal integration could be modeled by an interocular normalization model of population activity. These approach and model represent a necessary step in understanding how the binocular balance could be affected by deprivation such as short-term monocular occlusion.
Meeting abstract presented at VSS 2018