Resting state correlations are reduced between matched extrastriate regions in the blind (Watkins et al, 2012), and fine-scale connectivity is attenuated between the striate cortices (Butt et al, 2012 VSS). Here, we explore whether blindness alters the fine-scale connectivity and regional quadrantic (dorsal/ventral) relationship of extrastriate areas.
12 sighted and blind, age-matched participants were scanned (150 or 160 TRs BOLD fMRI, TR=3, 3mm voxels) under constant darkness. Within FreeSurfer template space each vertex was assigned a visual area (V2 and V3, dorsal and ventral) and retinotopic mapping value based on cortical surface anatomy (Benson et al., 2011 & 2013, VSS). We calculated a "quadrantic ratio", which relates the functional connectivity within the dorsal and ventral halves of visual areas to the connectivity between the two dorsal and two ventral regions. We next obtained fine-scale cross-correlation matrices (5595×5595 cells), which compare each extrastriate vertex to every other. These matrices were decomposed into a local-topography component (in which connectivity could be modeled as a Gaussian spread of correlation about two vertices that share retinotopic assignment) and a residual component.
The blind had decreased functional connectivity between right and left V2 and V3, and a significantly decreased "quadrantic ratio" within (p=0.009) and between hemispheres (p=0.037). While there were no group differences in fine-scale extrastriate connectivity as characterized by local topography (p=0.96), the residual matrices did differ between groups (p=0.026).
A surprising finding is that the local spread of correlation between topographically matched extrastriate points is unaltered in blindness. Vision is needed, however, to develop or maintain the link between the dorsal and ventral quandrants of a visual area and contributes to fine-scale correlations between visual areas that cannot be characterized by a local point-spread function.
Meeting abstract presented at VSS 2013