Abstract
Conventional wisdom suggests that frontal cortex is specialized for higher-level cognitive processes, such as spatial cognition and executive control. Recently, our laboratory demonstrated that caudolateral frontal cortex contains multiple, discrete, bilateral visual- and auditory-biased regions (Michalka 2015). Two regions, superior and inferior precentral sulcus (sPCS and iPCS), were recruited during visual vs. auditory attention. We also recently showed that visual- and auditory-biased regions in frontal cortex are revealed by contrasting functional connectivity to visual-biased regions in parietal cortex against auditory-biased regions in temporal cortex. Frontal regions exhibiting stronger connectivity with posterior visual regions not only include task-defined visual-biased regions, but also extend into additional portions of frontal cortex, such as rostral middle frontal gyrus (rMFG) (Tobyne, under review). Here, we further explore visuospatial task activation in frontal regions that are preferentially functionally connected to posterior visual structures. In individual subjects, we contrasted fMRI resting state functional connectivity between caudolateral frontal cortex and posterior visual and auditory seeds using a bivariate regression analysis. The resulting connectivity-defined visual-biased regions (sPCS, iPCS) were divided into two subdivisions: one that was recruited in the Michalka (visual vs. auditory) attention contrast ("original region") and an adjacent region which was not ("buddy region"). We examined activation within each subregion in multiple visuospatial fMRI paradigms, including multiple object tracking and visual short-term memory. Preliminary results suggest that activation during visuospatial tasks overlaps strongly with both "original" and "buddy" sPCS and iPCS regions, but rMFG is not as robustly recruited. Visual-biased regions identified by preferential functional connectivity appear to have different response profiles for different visuospatial tasks. We further suspect that the large sPCS and iPCS regions seen in many vision studies contain specialized sub-regions whose task response varies. Bivariate regression of functional connectivity profiles may be more powerful than traditional task activation in identifying specialized cortical structures.
Meeting abstract presented at VSS 2017