Training-induced recovery of motion perception after occipital stroke despite V1-V4 damage

Poster Presentation: Saturday, May 17, 2025, 2:45 – 6:45 pm, Banyan Breezeway
Session: Perceptual Organization: Neural mechanisms

Niki Lam1 (), Berkeley Fahrenthold1, Matthew Cavanaugh1, Bryan Redmond1, Jingyi Yang1, Duje Tadin1, Marisa Carrasco2, Noah Benson3, Elisha Merriam4, Krystel Huxlin1; 1University of Rochester, 2New York University, 3University of Washington, 4National Institute of Mental Health

Cortical blindness is often defined by damage, usually from stroke, to the primary visual cortex (V1) in the occipital lobe of the human brain. Here, we asked to what extent such strokes also affect extrastriate visual areas. We then assessed the relationship between damage, and the size and severity of behavioral defects, and the efficacy of visual motion discrimination recovery attainable with perceptual training. We analyzed data from 35 subacute occipital stroke patients (mean age/SD: 52.3/11.5 years; time post-stroke: 3.3/1.1 months; male/female: 23/12) enrolled in clinical trial NCT04798924 (ClinicalTrials.gov). Patients underwent Humphrey perimetry and eye-tracker-enforced psychophysical testing to measure fine direction discrimination (FDD) thresholds in their blind and intact hemifields. This was done before and after training on the FDD task with endogenous, feature-based attention (FBA) pre-cues for 6-11 months. Structural MRI scans of each brain were analyzed using the Benson atlas (Benson & Winawer, 2018) to estimate whether V1-V3, V3A/B and V4 were damaged. Damage exclusive to V1 was found in only 11% of patients; the rest exhibited damage to multiple, early visual areas. Patients with V1-only damage had smaller visual defects than those with combined V1 and extrastriate cortex damage [p = 0.01], although this was potentially biased by the small sample for V1-only damage. Importantly however, severity of the deficit and efficacy of FDD training did not differ between those with V1-only damage versus combined damage. Instead, training-induced improvements in FDD thresholds were directly proportional to the number of days trained. In conclusion, while occipital strokes rarely injured only V1, the number and type of extrastriate areas damaged did not correlate with pre-training preservation of blind-field motion discrimination abilities pre-training, or with the potential to recover motion discrimination with training. Our findings suggest that both preservation and recovery likely rely on visual pathways that bypass V1-V4.