Simulated low vision with young and old adults: How do they see?

Abstract

Purpose: The fovea plays a crucial role in human vision, but various conditions (e.g., Age-related Macular Degeneration) can cause an irreversible loss of central visual functioning. How well can people utilize peripheral visual function to perform visual tasks when their central visual field is impaired? Does age make a difference in how people adjust to a loss of central vision? How do compensatory eye movement patterns affect visual performance? To what extent can we dissociate attention from foveal vision? Method: We used a gaze-contingent technique to generate a simulated central scotoma to obscure observers' central visual field while they were performing a variety of computer-generated (MATLAB) visual tasks. Young and old observers with normal vision participated. Tasks included shape-from-texture identification, and discriminations of orientation, motion direction and velocity. Eye movements were recorded with an eye tracker. We measured task performance and many eye movement parameters with scotomas of different sizes. Results: We found rapid perceptual learning in both young and old groups. Observer's fixation position shifted as a function of scotoma size (only along X-axis). Visual performance in terms of dwell time, fixation duration, number of saccades, and saccade amplitude was significantly different between young and old observers. Old observers also had greater difficulties to initiate a first response when we used the largest scotoma (8 deg). Conclusion: Without a functioning central visual field, observers were still able to use the periphery to perform a variety of visual tasks that are usually performed with the intact fovea. Although it was more effortful to complete tasks, especially for old observers, young and old observers were both able to respond accurately to identification and discrimination tasks. This implies that the peripheral visual field can be trained relatively quickly to perform various qualitatively different tasks.

Footnotes

 This research was supported by grants from NSERC and FQRSC to MvG FQRNT to RZ.