Abstract
The traditional view of the interocular interactions in setting up ocular dominance in primate V1 is one of competition between the two eyes, at least during development if not also ongoing. A re-examination of ocular dominance (OD) data from optical imaging studies suggest an interocular process that seeks to maintain a set level of left/right eye opponency (an L-R signal) atop other ongoing cortical activity. Optical imaging of OD columns (ODCs) is performed by acquiring sets of V1 images during left eye (L), right eye (R) stimulation, and no stimulation (blank). The image data is analyzed to yield an ODC map with an L-R calculation. Invariably the OD (L-R) image yields a far "cleaner" map of the ODCs than the "single condition" maps provide (e.g. L-blank). Extracting a line profile across the ODCs in comparison to a line profile of the L or R maps confirms a smooth, nearly sinusoidal OD signal while the L or R profiles appear extremely noisy with the OD signal often buried. The calculation of L-R is equivalent to removing the "cocktail blank" signal, i.e. removing common mode cortical activity signal and other artifacts, beyond those removed with blank subtraction. These results imply an interocular process that provides a robust OD (L-R) signal despite substantial local variances in cortical activity. This notion is supported by short-term monocular deprivation (STMD) experiments in which a 1-2 hour monocular deprivation yields a disruption of interocular balance and monocular gains lasting an hour. Yet immediately following STMD, the OD(L-R) signal visualized by optical imaging "bounces back" to near normal in amplitude and form even while each eye's activity has not returned to baseline. The implication is that interocular/binocular V1 circuits strive for a robust OD(L-R) signal despite local and monocular perturbations, perhaps a by-product of maintaining proper interocular balance.
Meeting abstract presented at VSS 2018