Symmetry Produces Distinctive, Not Greater BOLD Activation in Object-Selective Cortex
Journal of vision(2018)
Abstract
Symmetry is a salient visual property. Previous fMRI studies have shown that symmetry activates object-selective cortex in humans, but this could be simply because symmetric objects evoked distinct patterns of activity. Here we investigated this issue by measuring brain activity using fMRI in human subjects while they viewed symmetric and asymmetric objects in two tasks. In the first task, subjects had to detect object repetitions in blocks of 12 serially presented items that were all symmetric or all asymmetric, but consisted of only silhouettes, only dot-patterns, or mixed silhouettes and dot-patterns. As in previous studies, we found that symmetric dot-patterns evoked greater activity compared to asymmetric dot-patterns in object-selective cortex. However, this advantage was abolished on comparing symmetric with asymmetric mixed objects. Thus, the stronger responses to symmetric objects observed in previous studies are likely due to BOLD signal adaptation. In the second task, subjects had to detect symmetric objects. We found that symmetric objects evoked similar activity in both early visual areas and in object-selective cortex. Yet, linear classifiers trained on multivoxel patterns were able to reliably classify objects as symmetric or asymmetric better in object-selective cortex compared to early visual cortex (average % accuracy = 69% and 64% respectively). Interestingly, distances of objects from the classifier boundary were correlated with symmetry detection time in both regions (r = -0.65 and -0.55 respectively, p < 0.0005). Taken together, our results show that symmetry perception in humans is mediated by distinctive, not greater activation of object-selective cortex. Meeting abstract presented at VSS 2018
MoreTranslated text
AI Read Science
Must-Reading Tree
Example
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined