Niki Lam¹ (), Berkeley K. Fahrenthold¹, Matthew R. Cavanaugh¹, Elizabeth Ginzburg¹, Neil Dogra¹, Maria Grothe¹, Bryan V. Redmond¹, Jingyi Yang¹, Duje Tadin¹, Marisa Carrasco², Krystel R. Huxlin¹; ¹University of Rochester, Rochester, NY, ²New York University, New York, NY

Occipital stroke causes postchiasmatic visual cortex damage and cortical blindness (CB), decreasing visual functioning and quality of life. Here, we asked whether motion discrimination training with feature-based attention (FBA) at a single, blind-field (BF) location for at least 6 months is sufficient to (1) increase performance at an untrained BF location, (2) improve functional field scores computed from Goldmann perimetry, and (3) improve quality of life as assessed with the National Eye Institute Visual Function Questionnaire (NEI-VFQ). Data were obtained from 39 subacute CB patients enrolled in clinical trial NCT04798924 (mean age/SD: 53.5/13.5 years; time post-stroke: 3.4/1.1 months; male/female: 25/14). Before and after training, patients completed eye-tracker–enforced psychophysical testing, perimetry, and the NEI-VFQ. Goldmann area for different isopters and functional field score (which heavily weighs the lower hemifield and central 10˚) were computed to estimate the extent of intact vision in each patient. The NEI-VFQs were manually scored and averaged to create a single composite score per person per test. At-home training involved performing 600 trials of a motion discrimination task with FBA and caused significant improvements at both trained and untrained BF locations. Additionally, training improved the functional field score, NEI-VFQ composite scores, and area of central isopter I2e, but not the total area of vision encompassed by the V4e isopter. Changes in motion-discrimination thresholds at trained BF locations did not predict changes in functional field score or NEI-VFQ composite score, but the number of sessions trained did, especially in the subacute period. In sum, localized BF motion discrimination training with FBA improved both local and global visual performance in CB patients, increasing the functional field of view and quality of life. Moreover, the amount of training during the subacute period predicted global visual function outcomes, suggesting that earlier and more frequent training is more effective.