Although we experience a unified perceptual world, the binding of features into integrated percepts is a complex neural process. Synesthesia, a condition in which one stimulus feature (e.g., letters) elicits the consistent experience of a second feature (e.g., color), has been argued to be an example of hyperbinding (Robertson, 2003), and it occurs within otherwise neurologically normal populations. Thus, identification of brain areas activated during synesthetic binding may lead to a better understanding of the neural processes involved in feature binding. Previous work has implicated a number of different parietal cortical areas in feature binding and synesthesia. However, within the parietal cortex, the spatial location of functional areas is highly variable across subjects. We therefore used region of interest (ROI) analyses to measure fMRI responses during a rapid serial visual presentation (RSVP) task that elicited color-grapheme perceptual bindings for letters in synesthetes but not control participants. Both groups of subjects also performed the RSVP task with non-letter symbols that did not elicit feature bindings in either group. Anatomical ROIs were defined for each subject, and each subject also completed a separate scanning session of retinotopic mapping to define topographically-organized areas within the intraparietal sulcus (IPS). We found that synesthetes, but not controls, exhibited greater activity for RSVP of letters compared to symbols in ventral parietal cortical regions. These ventral areas did not overlap with the more dorsally located topographic areas IPS0-5; activity in these dorsal areas did not differ between the participant groups. Furthermore, activity in ventral parietal cortex was predominantly in the right hemisphere, consistent with a right hemisphere bias for the ventral attention network. These results demonstrate that the parcellation of parietal cortex via retinotopic mapping and anatomical landmarks facilitates the identification of parietal regions involved in synesthetic binding.
Meeting abstract presented at VSS 2013