Abstract
A popular approach for assaying the binocular properties of cortical neurons employs monocular tests of ocular dominance to infer binocular function. However, the relationship between ocular dominance and binocular disparity—a property that is critical for stereopsis and can only be assessed with simultaneous stimulation of both eyes—remains unclear. Furthermore, it is not known whether a functional micro-architecture exists for binocular disparity. Using two-photon calcium imaging of individual cell bodies as a surrogate for spiking activity, we interleaved monocular stimuli and inter-ocular spatial phase disparity stimuli in single trials to explicitly determine the relationship between ocular dominance and binocular disparity for all neurons in 300 × 300 µm regions of layer 2–3 cat visual cortex (area 18). Our monoptic and dichoptic stimuli were 2 Hz drifting sinusoidal grating stimuli (50% contrast). We found that individual imaged sites often showed a clear compartmentalization for preferred inter-ocular spatial phase disparity. In addition, the precision of the disparity map was virtually identical at multiple depths. Using retinotopic mapping stimuli interleaved with each eye and the availability of more than a hundred reference cells per imaged site, vergence state was accurately determined. Trial-by-trial stability of retinotopy and preferred disparity, together with our results from varying spatial frequency with disparity, ruled out the possibility that the disparity maps were an artifact from our chosen stimulus set or use of an anesthetized preparation. When the full range of ocular dominance indices (1 through 7) were represented in a single 300 × 300 µm imaged site, single-cell resolution ocular dominance maps were often orthogonal to binocular disparity maps from the same site.
Supported by R01 EY017925, the Dana Foundation, and the Whitehall Foundation.