Steering through change: How aging eyes and stroke-induced vision loss affect driver control

Imaging system engineers could greatly benefit from understanding how biological systems have evolved to solve difficult imaging problems. In this thesis, we study how the human visual system has evolved to address the complex task of visually-guided steering, particularly as a person ages and experiences a prevalent type of vision loss from stroke called cortical blindness (CB) that results in the loss of one quarter to one half of the visual field. In a series of three studies, we test the hypothesis that the effects of CB on vision and steering extend beyond those that accompany healthy aging, and we predict that the effects cannot be accounted for by the common characterization of CB as an occlusion of visual information. Our approach involves a custom-made virtual reality steering task that facilitates the systematic exploration of how manipulating visual information affects both steering and gaze behavior. Studies 1 and 2 highlight the remarkable robustness of steering in the presence of aging visual systems and large visual impairments from CB, but they also characterize the widely variable behavior across CB individuals. In study 3, visually-healthy drivers were subjected to gaze-contingent masks and did not demonstrate the same biases as CB drivers. We concluded that the influence of CB on steering is more complicated than can be explained by a simple occlusion of visual information. Ultimately, we learned that the visual-motor strategies that facilitate steering are adaptable and resilient to the slow changes related to aging as well as the abrupt changes due to stroke. Our findings contribute knowledge to the scientific community about the role of vision for steering, and they also have the potential to inform future CB vision rehabilitation initiatives.