VSS 2015 Winners > Guido Maiello
Perceptual Consequences of Elongated Eyes
Guido Maiello1,2,3, William Harrison2, Fuensanta Vera-Diaz4, Peter Bex2; 1UCL Institute of Ophthalmology, University College London, London, UK, 2Department of Psychology, Northeastern University, Boston, MA, USA, 3Department of Informatics, Bioengineering, Robotics and System Engineering, University of Genoa, Genoa, Italy, 4New England College of Optometry, Boston, MA
Myopic eyes are elongated compared to the eyes of normally-sighted, emmetropic observers. This simple observation gives rise to an empirical question: what are the physiological and perceptual consequences of an elongated retinal surface? To address this question, we developed a geometric model of emmetropic and myopic retinae, based on magnetic resonance imaging (MRI) data [Atchison et al. (2005)], from which we derived psychophysically-testable predictions about visual function. We input range image data of natural scenes [Howe and Purves (2002)] to the geometric model to statistically estimate where in the visual periphery perception may be altered due to the different shapes of myopic and emmetropic eyes. The model predicts that central visual function should be similar for the two eye types, but myopic peripheral vision should differ regardless of optical correction. We tested this hypothesis by measuring the fall-off in contrast sensitivity with retinal eccentricity in emmetropes and best-corrected myopes. The full contrast sensitivity function (CSF) was assessed at 5, 10 and 15 degrees eccentricity using an adaptive testing procedure [Vul et al. (2010)]. Consistent with our model predictions, the area under the log CSF decreases in the periphery at a faster rate in best-corrected myopic observers than in emmetropes. Our modeling also revealed that a target at a given eccentricity projects onto a larger area of peripheral retinal for myopic than emmetropic eyes. This raises the possibility that crowding zones - the area over which features are integrated - may differ between eye types. We measured crowding zones at 5, 10 and 15 degrees of eccentricity using a 26 AFC letter identification task and found no significant differences between myopic and emmetropic observers. This suggests that crowding depends on spatial rather than retinal feature separation, which implies differences in the retino-cortical transformations in myopes and emmetropes.