Purchase this article with an account.
Andrew Carter Freedman, Jacob Achtemeier, Yihwa Baek, Gordon E Legge; Target Detection and Gaze Control with Reduced Acuity. Invest. Ophthalmol. Vis. Sci. 2018;59(9):3420.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To design visually accessible spaces for people with low vision, it is informative to understand their gaze behavior when encountering novel environments. The current study aimed to characterize how acuity restriction affects gaze behavior while subjects with normal and low vision searched for and identified targets during navigation. Because reduced acuity limits the amount of information processed in a glance, we hypothesized that reduced acuity would result in increased visual scanning behavior.
Subjects walked along an indoor path, guided by a rope line. They were instructed to look for objects, placed at unpredictable intervals left and right of the path, and to identify single letters posted on the objects. Performance was measured by time taken to complete the path and the proportion of targets detected. A Tobii head-mounted eye tracking system was used to measure gaze direction via a forward-facing video. Gaze direction was categorized as Left, Right, or On Path, see Figure 1. Recorded video was used to count the number of gaze transitions between categories during trials, quantifying scanning behavior. There were 8 low-vision subjects (mean acuity 1.09 logMAR, range .48 to 1.85 logMAR), and 8 normally sighted controls. Correlations between acuity, number of gaze transitions, and proportion of objects detected were assessed.
Number of gaze transitions and logMAR acuity were correlated in low-vision subjects, r =.42, p = .016, confirming that poorer acuity was associated with more scanning behavior. Number of gaze transitions was positively correlated with proportion of objects detected for controls, r = .59, p = .045 and negatively correlated for the low-vision group, r = -.63, p < .001. These results indicate that more gaze scanning meant more effective target search for controls, but more gaze scanning was associated with less effective target search for low-vision subjects.
Our hypothesis was confirmed that lower acuity in low vision is associated with more visual scanning. The more distributed gaze behavior is a compensation for the reduced visual information acquired in each glance. By comparison, when acuity is high (the normally sighted controls), more active scanning is associated with better search performance.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.
Figure 1: Gaze Direction Categories. The blue portion indicates the Left category, red indicates the On Path category, and orange indicates the Right category.
This PDF is available to Subscribers Only