Abstract
Spatial crowding corresponds to the degradation in visual acuity when a target is flanked by other patterns. We have recently found (Polat, Bonneh & Sagi, VSS03) that the spatial crowding observed in strabismic amblyopes is much larger than that found in ansiometropic amblyopes (N=61 patients) and is uncorrelated with the visual acuity for an isolated pattern. Moreover, about half of the crowding effect could be removed when the target and the flankers were presented in different colors (N=21). This suggests that at least part of the deficit is not due to lateral suppression by contrast masking, but rather is due to failure of selection in space. Accordingly, problems with top-down selection might be expected in the temporal domain as well. Here we measure “temporal crowding” and correlate it with spatial crowding in the same patients. We tested 7 strabismic (or combined) and 7 anisometropic amblyopes. Temporal Crowding (TC): an RSVP stream of digits (60 ms each, SOA=200,400) with target marked by reduced size (factor 0.7). Visual acuity was determined by the digit size leading to 50% correct identification. The TC effect was taken to be the difference between acuities for SOA=400 (un-crowded condition) and for SOA=200 (crowded condition). Spatial Crowding (SC): a computerized E-test with static stimuli, crowded and un-crowded (single digit) conditions. We found that acuities in RSVP for the “un-crowded” condition were correlated (R^2=0.48) with static visual acuities for an isolated pattern. The reduced SOA (200) of the crowded condition caused a degradation in acuity (temporal crowding effect) for some patients. The TC effect was correlated (R^2=0.72) with the spatial crowding effect. There was no correlation between TC and the spatial acuity for isolated pattern (R^2=0.001). We suggest that temporal crowding (TC) and spatial crowding (SC) reflect, at least in part, abnormal top-down attentional mechanisms which are responsible for both crowding effects.