Abstract
When a static gray surface is embedded in a temporally modulated in luminance surround, its brightness appears to change over time. The magnitude of the illusory brightness modulation is inversely related to the border contrast between the surface and the time-averaged surround luminance. Specifically, the perceived brightness modulation is largest when the contrast is zero, and gets progressively smaller as the contrast is increased. We used this inverse relationship between contrast and perceived (illusory) brightness in an fMRI experiment designed to identify the visual areas that underlie brightness perception.
Five subjects viewed static test surfaces of different contrast (0.0,0.2,0.5,1.0) created by changing the luminance of the surface (both increments and decrements) with respect to a time-averaged luminance of surround. The surround was modulated in luminance from 1050 to 1700 cd/m2 sinusoidally, at 1Hz. We analyzed retinotopic visual areas V1,V2, and V3 separately, using multiple regions-of-interest corresponding to different portions of the visual stimulus.If the region corresponding to the center of the test surface in a particular visual area is driven primarily by the border contrast (e.g.,Cornelissen et al,2006), we expected a progressive increase in the fMRI signal as we increased the contrast. However, if the visual area is primarily sensitive to perceived brightness we expected the opposite pattern - a progressive decrease in the fMRI signal with increasing contrast.
Increasing border contrast - thereby reducing the brightness induction - resulted in a progressive increase in the fMRI signal in V1 confirming this area’s sensitivity to luminance contrast. In V3, however, we observed the opposite pattern - an inverse relationship between contrast and the fMRI signal that approximated the perceived change in brightness. Area V2 showed an intermediate response. The data suggest that a progressive recoding of contrast to brightness information occurs in early retinotopic visual areas.
Supported by EY07031 from the NEI to MP and HM1582-05-C-0037 from National Geo-Spatial Intelligence Agency to SM