Minimum requirements for mobility in known environments and perceptual learning of this task in eccentric vision

Abstract

Abstract: : Purpose: Determination of minimum requirements for useful mobility and evaluation of learning effects when performing this task in eccentric vision. Methods: Mobility performance was assessed under conditions simulating artificial vision. Normal subjects wearing a portable artificial vision simulator were asked to complete an indoor course consisting of 6 randomly positioned obstacles. Five degrees of image resolution (17920, 1991, 498, 221, and 124 pixels) and 3 effective viewing angles (16.5°x11.6°, 33°x23.1°, and 66°x46.2°) were projected on a 10°x7° viewing area, stabilized at a fixed position in the visual field. First, the minimum requirements needed to reach optimal mobility performance were established using central vision. Then 3 naïve normal volunteers were trained to perform the same task using a viewing area stabilized at 15° of eccentricity in the lower visual field. Results: Using central vision, mobility performance decreased significantly (p<0.05) for image resolutions below 0.1 pixel/deg². There was no significant difference between the three viewing angles for greater image resolutions. The 33°x23.1° viewing angle tended, however, to be the best and subjects reported preferring it to the others. A viewing angle of 33°x23.1°, containing 498 pixels (0.7 pixel/deg²) was chosen for perceptual learning in eccentric vision. Subjects adapted quite rapidly to eccentric viewing. Within the first 10 training sessions error counts asymptoted. The time to accomplish the mobility task stabilized after about 40 sessions, reaching the same value than that measured using central vision. Conclusion: These data demonstrate that mobility in known environments requires much less image resolution than other tasks, like reading. This result is in agreement with clinical observations: low vision patients unable to read, can move surprisingly well in familiar settings. Performance in unknown, dynamic environments should still be investigated to get a more complete picture of minimum requirements for useful mobility.

Keywords: learning • space and scene perception • visual fields