Abstract
Introduction: Color sensitive neurons in the primate visual system are thought to fall into two distinct types: a minority that are color selective with little response to achromatic contrast, and a majority that respond well to both color and achromatic contrast. The functional roles of these two populations are unknown. Here we use an fMRI adaptation paradigm to investigate the selectivity of the human LGN, V1 and extrastriate cortical areas to color and achromatic contrast. Methods: The effect of adaptation was measured in a block design comparing adaptation and no-adaptation conditions. Adapting stimuli were high contrast isoluminant RG or achromatic sinewave counter-phasing rings (0.5cpd, 2Hz, 12 sec duration) and test stimuli were lower contrast RG or achromatic rings presented for 18 sec. We assume that cross adaptation of responses to chromatic and achromatic stimuli indicates a common neural substrate for both, whereas a lack of cross adaptation indicates selective neural regions. Regions of interest (ROIs; LGN, V1, V2, V3, VP, V3A, V4, hMT+) were independently localized using standard procedures. Functional data were analyzed using percent signal change and fixed effects GLM analyses contrasting responses following adaptation versus no adaptation. Results: We observed the greatest adaptation effect when the adaptor and test stimuli were similar: after chromatic adaptation BOLD responses were lower for chromatic than achromatic stimuli, but after achromatic adaptation responses were lower for achromatic than chromatic stimuli. This effect was found in all ROIs and provides evidence for selective adaptation. Notably, we also found ROI-dependent variation in sensitivity to adaptation. In particular, hMT+ was most affected by adaptation to achromatic contrast for all test stimuli and exhibited the least selectivity. Conclusion: The results support the functional presence of color selective neurons in the human visual system that can adapt independently to color contrast, with the likely exception of hMT+.
Meeting abstract presented at VSS 2014