Performance of low vision individuals when selecting a target with head-pointing in virtual reality
Purpose: To investigate psychophysically the ability of low vision individuals with central visual field loss (CFL) to perform a visually-guided pointing task in a virtual reality environment. Methods: Patients with CFL (n=25, ages = 67-90 years) and normally-sighted controls (n=26, ages = 67-85 years) had to select a target (2{\textdegree} diameter dot) with a head-contingent cursor (6{\textdegree} diameter reticle). Target selection occurred when target was validly pointed at for 1.5 seconds. Pointing was valid when target was inside an invisible pointer activation zone (PAZ) centered on reticle. Task difficulty was decreased by increasing PAZ diameter from 0.5{\textdegree} to 8{\textdegree}. Performance was assessed by measuring the time needed to select the target. The task was also performed with an array of three simultaneously-displayed cursors. Results: Selection times decreased (from 14.1 and 8.4 seconds for patients and controls respectively) with increasing PAZ diameter and reached a similar asymptote for both groups (1.4 seconds). The rate of this decrease was smaller for patients so that PAZ diameter needed for their best performance was much larger than PAZ diameter needed for controls' best performance (average: 3.48{\textdegree} vs 1.32{\textdegree}). In the three-reticle condition, both groups tended to use the cursor closer to the target. Conclusions: Patients with CFL are able to point at a 2{\textdegree} target thanks to head-pointing. Their performance can get close to controls' best performance by increasing PAZ size. Translational relevance: This research suggests guidelines to improve the accessibility of visually-guided pointing tools for human-machine interfaces designed for low vision individuals.