Jonathan K. Doyon^1,2, Madeleine Heynen^1,2, Wei Hau Lew^1,2, Claire E. Manley^1,2, Alex D. Hwang^1,2, Jae-Hyun Jung^1,2, Lotfi B. Merabet^1,2; ¹Massachusetts Eye and Ear, ²Harvard Medical School

Cerebral visual impairment (CVI) is a brain-based visual disorder and the leading cause of pediatric visual impairment in the developed world. Typically, CVI arises as a result of early neurological damage and maldevelopment of retrochiasmal visual processing pathways. Individuals with CVI often report an inability to tolerate dense and highly dynamic visual environments (e.g., crowded shopping malls). Consequently, safely navigating these environments becomes nontrivial, requiring attentionally demanding adaptive strategies. To investigate how individuals with CVI safely negotiate such environments, we asked subjects to complete a collision-avoidance task in an immersive virtual shopping mall displayed in a Meta Quest Pro head-mounted-display with integrated eye tracking. Two CVI (25.5 years ±0.71) and 4 healthy control subjects (27.25 years ±3.40) walked unrestricted through crowds of virtual pedestrians (densities of 2, 10, or 20 people) while detecting (via button press) and avoiding potential collisions (e.g., by altering walking speed or path) from target pedestrians approaching from various bearing angles (±20°, ±40°, or ±60°). CVI subjects performed comparably to controls in terms of collision detections, response times, and successful avoidances. However, CVI subjects showed altered gaze profiles with greater time-to-first-fixation of the target (+440ms, β=0.76, p=0.02), greater fixation durations (+588ms, β=0.75, p<0.001), and fewer fixations (-0.67 fixations, β=14.91, p=0.003). CVI subjects also walked slower than controls (-0.05m/s, β=-0.07, p=0.049). For both groups, crowd density slowed response times (+561ms, β=0.29, p<0.001) and resulted in greater time-to-first-fixation (+438ms, β=0.27, p=0.003). These results suggest that the CVI subjects have developed strategies to safely navigate crowded spaces, such as more conservative walking and different gaze behaviors. We suspect that these alterations result from an impaired ability to balance the stimulus-driven and goal-directed aspects of the task, evidenced by the CVI subjects being slower to fixate the target, and maintaining longer fixations on the target to ensure safe passage.