Abstract
Connective field modeling (Haak 2013) assigns retinotopy to cortical points by modeling the connective field (ccRF) within early cortical areas, using data collected with retinotopic mapping stimuli (e.g. bars, wedges, rings) while subjects maintain fixation. Here we test if the ccRF approach yields useful retinotopic maps from data collected while subjects view a naturalistic movie without fixation. The retinotopic organization in V1 was assumed from a cortical surface template (Benson 2012; 2014) then projected to extrastriate cortex and subcortical visual structures (LGN, SC) using connective field modeling. Three subjects were studied at 7T (voxels=1.5mm3; TR/TE=2000/25ms) during 30 minutes of drifting bar stimuli with central fixation, and 30 minutes of free-viewing of Pixar shorts. Retinotopy of each cortical and subcortical voxel was determined using connective field modeling, maximizing time-course correlation with a V1 patch. Moving dot and flashing checkerboard stimuli were used to localize subcortical structures. Cortical retinotopic maps generated from the movie data were comparable to those obtained from traditional retinotopic mapping methods, with polar angle reversals extending beyond hV4 and V3a. The freely viewed stimuli produced a greater extent of mapped cortex at greater eccentricities. Subcortical ccRF analyses identified the LGN reliably, and the SC more weakly. Retinotopic assignment within the LGN was readily seen when ccRF approaches were applied to data collected with fixation and drifting bars, less for free movie viewing. It is unclear if this is a stimulus or fixation effect. Furthermore, we find that connective field size is not significantly different in the free-viewing runs compared to drifting bar stimuli across visual areas. These results indicate that retinotopic mapping in extrastriate cortex can be accomplished using connective field modeling and naturalistic stimuli without fixation, but that fixation and/or traditional mapping stimuli may be required to identify and map subcortical visual structures.
Meeting abstract presented at VSS 2015