DESCRIPTION (provided by applicant): The goal of this proposal is to gain a better understanding of the information processing and decision strategies that underlie eye movement planning in both the normal and diseased state. In patients with age-related macular degeneration (AMD), central areas of the retina are damaged, creating a large blind spot that forces them to rely solely on residual vision in the periphery. Rehabilitation outcomes for these patients can be successful, but are often inconsistent. Despite similar retinopathies, some patients learn to use their residual vision more effectively than others. We have developed an information-theoretic model and experimental paradigm which will allow us to objectively measure human scanning efficiency. The development of the model has naturally motivated fundamental experimental questions about eye movements and neural decision making. The answers to these questions will be used to refine the model and enhance our understanding of the system in general. We will then apply the model framework to investigate differences in eye movement behavior between AMD patients and normally-sighted individuals. The interplay of model development and experimental investigation will significantly increase our knowledge of how humans use prior knowledge and task demands to direct their gaze, and how new visual information is incorporated into an eye movement plan. The results will have broad relevance to understanding neural decision making in general. Relevance to Public Health. The application of the model to a clinical population will bring much-needed objective measures to understanding the extent of impairment in individuals with AMD. With this understanding comes great potential for improving rehabilitation training strategies that will enhance the quality of life for these patients and their families.