Abstract
Purpose:
In attempts to better understand the effects of visual feedback on fixation stability in amblyopes, previous studies have included an ‘open loop’ test condition, in which the eye movements of the non-viewing (covered) eye were assessed. But interpretation of the results has been hampered by the incomplete knowledge of the link between eye movements of the two eyes. Here we study fixation stability in a ‘pseudo-open loop’ condition with reduced visual feedback to the viewing eye(s).
Methods:
Seventeen children with amblyopia (age 5-15 years; visual acuity 0.2-1.2 logMAR) and 10 age-similar normal controls participated. Fixational eye movements were recorded at 500 Hz under monocular (amblyopic eye) and binocular viewing conditions using two fixation targets: a small 0.3°-diameter dot and a 2-dimensional Gaussian profile (σ=1.25°), the latter appearing as a blurred disk providing only minimal visual feedback. Fixation stability was studied in terms of the bivariate contour ellipse area (BCEA) and the mean amplitude and rate of microsaccades.
Results:
In controls, fixation stability was best when binocularly viewing the dot (BCEA, -0.52±0.71 log deg2) and poorest when monocularly viewing the Gaussian profile (+0.60±1.09 log deg2). Similarly, in the amblyopic children, fixation stability was best when binocularly viewing the dot (BCEA, -0.17±0.53 log deg2) and poorest when viewing the Gaussian profile with only the amblyopic eye (+0.44±0.47 log deg2). However, the deterioration of fixation stability under reduced visual feedback was smallest in the amblyopes when viewing with the amblyopic eye. Group differences in BCEA were larger when viewing the dot than when viewing the Gaussian profile (2-way ANOVA, P<0.03). For both controls and amblyopes, the deterioration of fixation stability under reduced visual feedback was associated with increases in amplitude but not rate of microsaccades (P<0.016).
Conclusions:
Amblyopes showed a smaller-than-normal decrease in fixation stability when visual feedback was reduced in the ‘pseudo-open loop’ condition, consistent with a weakened input of visual feedback into the fixation control mechanism. This effect may hold promise as an objective ocular motor measure for the detection of amblyopia.