Incomplete letter recognition is limited by cortical and not optical factors: Simulating the deficits of visual-led dementia in healthy adults

Poster Presentation: Saturday, May 17, 2025, 8:30 am – 12:30 pm, Pavilion
Session: Spatial Vision: Clinical

Zien Huang1 (), Tessa M. Dekker1, Sebastian J. Crutch1, Keir X. X. Yong1, John A. Greenwood1; 1University College London

The Graded Incomplete Letters Test (GILT) was recently developed to detect visual symptoms in dementia, including Posterior Cortical Atrophy (PCA; a neurodegenerative syndrome predominantly affecting visual cortex). The GILT measures the threshold for recognising letters degraded by removing pixelated letter sections (decreasing ‘completeness’). Thresholds are strongly elevated in PCA patients relative to typical adults (Yong et al, 2024), though it is unclear why PCA patients struggle on the task. We aimed to distinguish the role of cortical factors associated with PCA (elevated crowding and impaired feature integration) from disruptions by age-related optical factors (e.g. blur or low contrast through glaucoma) that could be mistaken for PCA-related visual loss by simulating these effects in typical adults (n=6). To examine optical factors, we applied blur and lowered contrast separately to incomplete letter stimuli, with participants required to identify one of 12 uppercase letters presented foveally on each trial. Stimuli were degraded at different completeness levels using QUEST. Without optical factors applied, thresholds averaged ~5% completeness. With blur or low contrast, there was very little effect until blur/contrast approached detection/visibility thresholds, where small elevations to 8% completeness were evident. These deficits do not reach the levels seen in PCA (median: 47% completeness). To examine cortical factors, we simulated elevated crowding (impaired object recognition in clutter) by moving stimuli into peripheral vision, where crowding is high in typical adults. We simulated issues in feature integration by varying pixel size to alter the distribution of letter degradation (limiting the spatial integration of letter features), or by degrading letters dynamically with limited-lifetime pixels (limiting temporal integration). The combination of elevated crowding and feature integration impairments increased thresholds as high as 40% (with dynamic presentation in peripheral vision), comparable with average PCA deficits. Poor incomplete letter recognition is therefore more strongly associated with cortical than optical factors.