Abstract
Purpose: Most assessments of interocular balance or suppression in amblyopia use rivalrous stimuli, which do not reflect naturalistic vision. We developed a task in which observers continuously report the perceived contrast of dichoptic grating stimuli, identical to each eye except that the stimulus is contrast-modulated independently over time. Methods: Using a mirror stereoscope, observers (n = 16 amblyopia, n = 12 controls) viewed a 2-cpd Gabor that rotated slowly, while modulating in contrast at independent rates in each eye. Observers used a joystick lever to match the perceived contrast over time. We fit a model that predicts the time-course of perceived contrast from the time-course of the contrasts in each eye. Our model parameterized attenuation as a reduction of stimulus input to the amblyopic eye and suppression as interocular normalization. We also included a dichoptic letter chart to measure balance point (the relative strength of each eye for a rivalrous stimulus). Results: Model fits show that input to the amblyopic eye was significantly more attenuated in amblyopia compared to controls (p < 0.001). Both suppression of the amblyopic by the fellow eye, and suppression of the fellow eye by the amblyopic eye, were not significantly different in the amblyopia group compared to controls (p = 0.50 and 0.74, respectively). Balance point was significantly correlated with attenuation (r = 0.48, p = 0.002), but not suppression. Conclusion: Our dynamic contrast estimation task provides estimates of attenuation and suppression under naturalistic viewing conditions. Perhaps unsurprisingly, under these conditions, suppression plays a much weaker role in amblyopia than has been observed in previous studies using rivalrous stimuli. These data also raise the possibility that the severity of suppression may not be tightly linked to amblyopic deficits in all observers, reflecting independent targets for treatment.