The role of the parietal cortex in feature binding in visual search

Abstract

 

The ability to combine simple visual features in the environment to distinguish individual objects is an essential aspect of visual perception. This process, called feature binding, has previously been described as a mapping of multiple features to a location in space (Treisman, 2006). Areas of the parietal cortex have been shown to have specificity for different forms of visual attention related to stimulus processing. In particular, dorsal parietal attention areas contain retinotopic maps for spatial attention, while ventral parietal attention areas are lateralized to the right hemisphere and have not been shown to have this kind of location specificity in their responses. Evidence from both patient and TMS studies suggests a role for the parietal cortex in feature binding. Using a within-subject fMRI design, we compared cortical regions involved in two types of visual search. Subjects detected a target defined by a single feature (color pop-out search) on some trials, while on other trials the target was defined by a conjunction of two features (color-shape conjunction search). Task difficulty was matched for the two types of search by manipulating stimulus-mask onset asynchrony. We used retinotopic mapping to define early visual and dorsal parietal (IPS0-5) areas and found that parietal cortical activity distinguished between feature and conjunction search only in areas outside of the topographically-organized dorsal attention network. This activity was consistently centered within the anatomically-defined angular gyrus and was predominantly in the right hemisphere, consistent with the previously reported right-hemisphere bias for ventral parietal networks. We therefore support other evidence that the ventral parietal cortex plays an important role in feature binding, while the retinotopic dorsal areas of the parietal lobes show less involvement.

 

Meeting abstract presented at VSS 2013