Abstract
Naturally occurring cases of extreme variation in visual system anatomy and neurophysiology, as in albinism, can aid identification of key factors responsible for brain-related vision deficits. To this end, we have quantitatively characterized retinal cone density, cortical magnification and population receptive field structure in 5 subjects with albinism, ultimately to develop a quantitative neurophysiological model predictive of visual dysfunction. In albinism, foveal cone density and Vernier acuity are reduced relative to healthy controls at all eccentricities from fovea to periphery (20°), yet acuity does not directly scale with cone density over this range. Accordingly, cortical magnification at/near the fovea is reduced relative to controls yet can be higher in the periphery than in controls (though this is markedly variable across subjects). fMRI-based cortical retinotopy and voxel receptive field structure are also markedly aberrant due to mis-routing of retinal afferents at the optic chiasm. This results in partially superimposed representations of the left and right visual fields within the same hemisphere and to a significant percentage (14%) of voxels with “dual” population receptive fields in V1–V3. The latter respond to visual stimuli at 2 separate locations, previously thought to be mirror symmetric across the vertical meridian. However, our pRF modeling of 2200 dual pRF voxels revealed that 32% were at least 45° off mirror symmetric, some with components in the same visual hemifield and/or straddling the horizontal meridian (incidence varying considerably across albinism subjects). This suggests that retino-cortical miswiring can be more complex than a simple decussation artifact at the chiasm. Together these results reveal that anatomy and neurophysiology are more variable and complex in albinism than previously appreciated. This may correlate with genetic variation within the albinism population and provides a rich source of “natural experiments” for exploring brain-related structure-function-deficit relationships.
Acknowledgement: NIH grants TL1TR001437, T32GM080202, T32EY014537, P30EY001931, and R01EY024969