Impact of Crowding in Amblyopia

A new Levi Lab study explores the impact of crowding on visual appearance and performance in amblyopia.

What does the world look like through an amblyopic eye? A new study from the Levi Lab explores how people with amblyopia perceive visual stimuli and why those differences may extend beyond reduced visual acuity. Published in Vision Research, the paper was led by Postdoctoral Researcher Angela Tomaz, with co-authors Wolf Harmening, Dennis Levi, and Bilge Sayim.

Abstract

Individuals with amblyopia have been reported to perceive stimuli as distorted with the amblyopic eye, and to have stronger foveal crowding than controls without amblyopia. Previous studies on visual appearance in amblyopia often lacked a control group and did not adjust stimulus size to account for visual acuity deficits.

Here, we investigated whether observers with amblyopia perceived stimuli as more distorted and whether foveal crowding influenced visual appearance more strongly than in controls, while controlling for visual acuity differences. Isolated and flanked target letters were presented foveally to the dominant/fellow and non-dominant/amblyopic eye. Stimuli were scaled relative to each eye's acuity threshold and presented at 1.0× and 1.5× threshold size. Observers replicated target appearance by selecting squares on a novel grid interface.

Results showed that accuracy, structural similarity (SSIM) and perceived similarity (PSS) were lower for observers with amblyopia compared to controls (Group), for small compared to large stimuli (Size), and for flanked compared to isolated targets (Crowding). Interestingly, a significant interaction between Group and Crowding was observed only for accuracy and SSIM, but not for PSS. These results showed group differences despite controlling for acuity, suggesting that factors beyond visual acuity, such as stronger foveal crowding, contribute to visual appearance differences in amblyopia compared to controls.

About the Image

Mean pixel selection in the responses of both groups (control, amblyopia) for each target letter in both sizes (1.0 and 1.5 x threshold), eye (DE/FE, NDE/AE) and crowding conditions (Isolated, Flanked). 

Read the Paper