Abstract
It is well known that the control of steering (e.g., when driving) is reliant on visual information from optic flow (Kountouriotis et al. 2016). Because optic flow is spatially correlated and accurate heading judgments can be made using a sparse and partial flow field (Warren and Kurtz 1992), it is surprising that drivers with cortical blindness (CB) across ¼ to ½ of their visual field demonstrate more variable lane positioning than their visually intact counterparts (Bowers et al. 2010). We hypothesized that this deficit arises because residual noise introduced in the “blind” field affects optic flow processing in service of steering. To test this hypothesis, we analyzed steering behavior in 10 CB drivers and 5 visually-intact controls immersed in a virtual reality steering task. Participants were asked to maintain a center-lane position while traveling at 19 m/s on a procedurally generated one lane road. Turn direction (left/right) and turn radius (35, 55, or 75 m) were manipulated. Additionally, optic flow density was indirectly manipulated through variation in environmental texture density (low, medium, high). Analysis of the average distance from the inner road edge revealed that all CB drivers were biased away from their blind side, but only controls and those with right-sided deficits decreased their distance to the inner road edge on medium and high optic flow density trials. The difference between these groups and the steering behavior of left-sided CBs, who showed no impact of optic flow, could not be attributed to age differences, time since stroke, or sparing in the central 10° of the visual field. Our results suggest that left-sided CBs place less weight on optic flow than right-sided CBs and controls. Preliminary analysis of gaze data suggests the insensitivity to variations in optic flow might also be attributed to compensatory gaze behavior.