Abstract
The spatial extent of crowding is elongated toward the anatomical fovea. Nandy & Tjan (2012) argued that this anisotropy is the result of acquiring image statistics under saccadic eye movements, which included uncompensated image movements caused by saccades. This theory predicts that the shape of the crowding zone may change if saccades are no longer referenced to the fovea, as in patients who have adopted a preferred retinal locus (PRL) for fixation after central field loss. Recently, our group has developed a technique to induce a PRL in normally sighted individuals using a simulated central scotoma (Kwon, Nandy & Tjan, 2013). Here we tested whether this experimentally induced PRL could lead to the predicted changes in crowding zones. Seven observers participated in 20-hours of PRL training, using the Kwon et al. (2013) method. We measured the two-dimensional crowding zone before and after training at 2 locations: the prescribed PRL location in the upper left visual field (eccentricity = 6.5°, 30° left of the vertical meridian) and a peri-PRL location positioned symmetrically in the right upper field. Under the condition of simulated central scotoma and after 20 hours of training, all seven observers developed a PRL at the prescribed location, with their saccades re-referenced to the PRL. We have thus replicated the findings of Kwon et al. (2013) and extended them to a prescribed (as opposed to spontaneously emerging) PRL. PRL training had a main effect in reducing crowding along the radial direction (p< 0.05). This result is consistent with a rounding of the crowding zone at the PRL location, as observed in AMD patients by Chung (2013) and predicted by Nandy & Tjan. However, evidence for the predicted reorientation of the crowding zone in the peri-PRL location was equivocal. Experimentally induced PRLs had an effect on crowding, albeit limited.
Meeting abstract presented at VSS 2016