Abstract
Abstract: :
Purpose: An augmented vision system consisting of an optical see–through HMD system that superimposes a minified edge image of the real world has been proposed to expand the visual fields of tunnel vision (TV) patients. We compared performance in visual search with and without the system to determine how much improvement in search performance could be achieved with the vision multiplexing design, and then how the system design should be improved. Methods: 12 subjects with TV (7–16 deg) searched for targets outside their residual fields on a blank background. Three subjects carried out search over a 90x74 deg area, and nine subjects over a 66x52 deg area. Eye and head movements were recorded using a head mounted eye tracker and a motion tracking system. Recording segments from the time subjects started to move their gaze from a central fixation point to the time they fixated the targets were extracted for analysis. Tasks consisted of 20 trials for each of three eccentricities (20, 27, and 35 deg for the large area; 15, 22, and 29 deg for the small area), which were all within the minified display. Results: For the large area, the search time with HMD was about half the time without it. For the small area, the improvement was not as large. Search time, compensated for habitual motion speed, was highly correlated to visual field size (R2>0.45 for the 22 and 29 deg eccentricities). Plots of search time as a function of field size without HMD and with HMD crossed each other, suggesting that subjects with fields >10 deg could still find targets faster with HMD than without it for the small search area. Varying levels of disorientation were reported by subjects and were suggested as well from the analysis of eye movement patterns. Conclusions: The augmented–vision HMD does help TV subjects with search over wide areas, with which they would otherwise have great difficulty. The relatively small benefit of the system for search over small areas might have been due to subjects’ ability to search such areas reasonably effectively without an aid. Inexperience with the system might have contributed to the disorientation. To derive greater benefit from the system, patients will require training; orientation marks in the display might also be helpful. Supported by NIH grant EY12890.
Keywords: visual search • visual fields • low vision