Abstract
Purpose: :
Patients with central visual field loss due to diseases such as macular degeneration not only have difficulties seeing details but also spend a longer time on a task even if acuity is not an issue. How is time spent when foveal vision is made unavailable? We studied normal subject’s behavior in search tasks when a simulated central scotoma blocked foveal vision.
Methods: :
A head-mount, 500Hz eye tracker was used to generate a gaze-contingent display in which a 6 or 10 deg square area at the gaze location was painted to the display background color, effectively producing a central scotoma. Search performance in 3 tasks, a large square target among smaller square distracters, a red vertical bar target among red horizontal or green vertical bar distracters, and an "O" among "C" distracters, were studied in 5 normal subjects. The subject pressed a key if the target was present and pressed another key if the target was absent.
Results: :
Two eye movement behaviors were quantified for each correct search trial. A flashback was two consecutive saccades that left and then returned to the same area. A regression was a return to a searched area after visiting multiple locations. The following factors significantly affected the number of flashbacks and regressions (F&R). A larger scotoma resulted in more F&R than a smaller scotoma. A larger search set size (32 items) resulted in more F&R than a smaller set size (1 or 8 items). Target-absent trials produced more F&R than target-present trials. The most difficult search task (O in Cs) produced more F&R than easier tasks (size feature search or color/orientation conjunction search). Natural viewing or a simulated tunnel vision did not produce any F&R. The number of F&R was reduced after the subject had practiced the search tasks.
Conclusions: :
When the use of foveal vision was unattainable, frequent revisits to searched locations were observed, especially in more difficult search tasks. This behavior might not stem from impaired acuity. Difficulty with attention deployment, crowding effect and lack of confidence in peripheral viewing might have contributed to this behavior.
Keywords: visual search • low vision • eye movements