Abstract
Abstract: :
Purpose: Amblyopes are deficient at detecting second order structure (Wong et al. 2001). We wondered if they are also deficient at integrating second order structure since previously (ARVO, 2002) we have shown that amblyopes can integrate first order structure as well as normals. Methods: 5 amblyopic and 3 normal observers were recruited for this experiment. The visual acuity of the amblyopic observers varied between 20/40 to 20/200. Arrays of 16 second-order (contrast-modulated) Gabor micro-patterns were presented to the observers. Observers judged whether the mean orientation of the array was tilted to right or left of vertical. Prior to this, we matched the visibility of a single Gabor for amblyopic eyes (AE) and non-amblyopic eyes (NAE) by first measuring the detection threshold for the carrier (2D 1/f noise) and then using a horizontal/vertical discrimination task to measure the detectability of the modulation. Finally, by presenting one oriented second-order Gabor at 1, 1.5 and 2 times its threshold modulation depth, we equated orientation discrimination performance for both eyes of each observer. Observers did all tasks monocularly. Results: Two severe amblyopes showed significant deficiencies in detecting the carrier when using their AE. All amblyopic observers needed higher modulation depths than normals to detect isolated second-order stimuli with their AE and NAE. For the orientation integration task, AE, NAE and normal eyes exhibited the same performance levels and were found to have similar levels of high internal noise and low sampling efficiency. Conclusions: Although AE are deficient in terms of first- and second-order inputs, if the visibility of the carrier and modulator are equated, the integration of second order orientation information is normal.
Keywords: amblyopia • spatial vision • texture