Purchase this article with an account.
K Knoblauch, L Piettre, J Warnking, C Delon-Martin, C Segebarth, M Dojat; Visual Areas Involved in Color Scission . Invest. Ophthalmol. Vis. Sci. 2002;43(13):3984.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: The phenomenon of color transparency requires that the visual system assign two colors to the same region of the retinal image, one to a filter and a second to the surface seen through it. The visual system integrates local color differences to extract the global color of the filter, a phenomenon called color scission. We manipulated local color differences while using Event-Related (ER) fMRI to locate candidate visual areas involved in this task. Methods: Twelve subjects with normal vision were examined (1.5 T, EPI, TR/TE/flip=2500ms/50ms/85deg, 28 transverse slices, voxel size=4x4x5mm^3). The ER paradigm was composed of a pseudo-random presentation of 99 isoluminant images, each lasting 0.5s and followed by a rest screen (2s). Three types of images were employed: 1) Transparency, in which a translation was applied to the color coordinates of 10 central surfaces in a square region of a Mondrian; 2) Shearing, in which the direction of the translation was inverted for half of the surfaces; and 3) A null event used as a baseline. During recordings, subjects responded as to whether the image included a transparent region or not. Random effect group analysis was performed using SPM99 software. For comparison, subjects were also tested with achromatic and isoluminant Mondrians. Results: Subjects classified images with translations as containing a transparent region and shearing not. Transparency produced activations in frontal areas and in a medio-anterior part of ventral occipital cortex (left: [-16,-42,11], Zmax=2.86, right:[36,-50,7], Zmax=3.19, p<0.01 uncorrected) that did not overlap those areas differentially activated by isoluminant and achromatic Mondrians (also, ventral occipital: [±29, 67, -11] , Zmax=4.4, p<0.001 uncorrected). Interestingly, the shearing condition activated superior occipital gyrus ([27 74 29], Zmax=2.89, p<0.01 uncorrected). Conclusion: Our results indicate that color scission activates a different region than that by color alone. The parieto-occipital activation during the shearing condition could result from the subjects' task, if we suppose that perceptual integration leading to transparency produces 'pop-out' and that classification of stimuli as non-transparent requires focal attention.
This PDF is available to Subscribers Only