Abstract
Parsing a visual scene requires the differentiation of foreground and background figures, which may be distinguished by luminance contrast edges. However, when two objects are colored identically, their contours can be inferred. Early visual cortex responds to illusory contours (modal completion) where abutting lines or collinear edges imply the presence of an occluding surface. Further, it has been shown that early visual cortex signals also encode occluded objects behind a surface (amodal completion; Ban et al., 2013). Here we used functional magnetic resonance imaging (fMRI) and population receptive field (pRF) analysis in normal healthy volunteers (n=7) to map the visual cortex with bar stimuli that traverse the visual field. Stimuli were either defined by modal or amodal completion or comprised a subtle luminance contrast. Meanwhile, participants performed an orthogonal detection task at fixation. All three conditions produced reliable retinotopic signals within V1-V3. Compared to a separate pRF mapping experiment using conventional stimuli, signal strength was considerably weaker and pRF size larger for all conditions. Importantly, we found no significant differences in activation or pRF size between the three experimental conditions. Our findings suggest the possibility that despite the orthogonal fixation task these conditions merely measured the topographic signature of spatial attention to the bar stimulus rather than any specific activation related to illusory contours or occlusion. This points towards a need for careful experimental design in future studies that seek to infer specific stimulus processing from retinotopic signals.
Meeting abstract presented at VSS 2017