Altered Perception of the Bistable Motion Quartet in Albinism

Abstract

Purpose: Perception of the motion quartet (MQ) alternates between horizontal and vertical motion, with a bias toward vertical motion. This vertical bias has been explained by the dominance of intrahemispheric processing. In albinism, each hemisphere receives input from both visual hemifields owing to enhanced crossing of the optic nerves at the optic chiasm. This might affect the perception of the ambiguous MQ and particularly the vertical bias.

Methods: The effect of optic nerve misrouting in persons with albinism and nystagmus (PWA, n = 14) on motion perception for MQ was compared with healthy controls (HC; n = 11) and with persons with nystagmus in the absence of optic nerve misrouting (PWN; n = 12). We varied the ratio of horizontal and vertical distances of MQ dots (aspect ratio [AR]) between 0.75 and 1.25 and compared the percentages of horizontal and vertical motion percepts as a function of AR between groups.

Results: For HC, the probability of vertical motion perception increased as a sigmoid function with increasing AR exhibiting the expected vertical percept bias (mean, 58%; median, 54%; vertical motion percepts). PWA showed a surprisingly strong horizontal bias independent of the AR with a mean of 11% (median, 10%) vertical motion percepts. The PWN was in between PWA and HC, with a mean of 34% (median, 47%) vertical perception. Nystagmus alone is unlikely to explain this pattern of results because PWA and PWN had comparable fixation stabilities.

Conclusions: The strong horizontal bias observed in PWA and PWN might partly result from the horizontal nystagmus. The even stronger horizontal bias in PWA indicates that the intrahemispherical corepresentation of both visual hemifields may play an additional role. The altered perception of the MQ in PWA opens opportunities to (i) understand the interplay of stability and plasticity in altered visual pathway conditions and (ii) identify visual pathway abnormalities with a perception-based test using the MQ.

Figure 1.

Testing paradigm of motion perception for the MQ display. (A) Schematic of the test sequence and (B) percepts to be reported. See Methods for details and Supplementary Table S1 for the specific geometry of the stimuli including the different ARs applied. The red dashed rectangle indicates one observation sequence.

Figure 2.

Results of the generalized linear mixed-effect binary logistic regression model displaying the predicted probabilities (mean ± SEM) of vertical motion perception across varied AR and two different ways to vary the AR (vertical or horizontal dot distance constant). Red (horizontal) dashed line indicates chance level, black (vertical) dotted line indicates equal vertical and horizontal dot distances of the MQ stimulus, AR = 1. Larger than 50% vertical motion percepts at AR = 1 in the HC group indicates the vertical bias of the MQ, as known from the literature. In contrast, a strong bias toward horizontal motion percepts can be observed in the two patient groups. Analysis and statistics were based on only 6 (out of 9) horizontal/vertical AR around 1: 0.75–1.25. Data points not included in the analysis (AR 0.45–0.65) are grayed out.

Figure 3.

Comparison of percentage of vertical motion perception averaged for each subject across all trials. ω², partial omega squared effect size; CI, confidence interval. The median value is given with untransformed % tag. Analysis and statistics were based on only six horizontal/vertical AR around 1: 0.75–1.25. Small filled circles indicate individual data points. The large filled circles indicate within-group means.

Figure 4.

Correlation between vertical motion perception and the VEP-based optic nerve misrouting index of checksize 0.5° (interocular correlation of the interhemispheric VEP difference detailed in Methods). Dashed line indicates cutoff for misrouting versus no-misrouting coefficients (<0.0 [left], misrouting; >0.0 [right], no misrouting). Black regression line (behind gray dashed) and its confidence interval are shown for the VEP checksize 0.5°. For comparison purposes, dark gray dashed line for VEP checksize 1.0° and dotted line for VEP checksize 2° are also shown in the figure.