Purchase this article with an account.
Michael P. Barry, Gislin Dagnelie, H. Christiaan Stronks; Successful Target Identification with the Argus® II Retinal Prosthesis. Invest. Ophthalmol. Vis. Sci. 2012;53(14):5509.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To demonstrate the ability of blind subjects, using prosthetic vision, to locate and count targets toward a hand-eye coordination task.
Four subjects blind from retinitis pigmentosa and using and Argus II retinal prosthesis viewed black sheets containing the white frame outline of a square 18 centimeters wide, placed on a reading stand at 40 cm distance. Within the outline, there was any number from 1 to 8 solid white square targets, of 2.25 x 2.25 cm², in semi-random configurations. Subjects selected a preferred outline width. They were told the range of target numbers that could be presented inside the outline and were instructed to count the number of targets on each sheet presented. Sheets with different numbers of targets were presented in random order, in blocks of 8 to 12 sheets per session. Subjects received training in the lab and were allowed to practice at home. Outcomes recorded were percent correct, number of squares under-/over-counted, and time spent.
Three of the 4 subjects demonstrated learning with decreasing time per trial spent in successive blocks. For each of these subjects, under-counts outnumbered over-counts, but over-counts were larger in magnitude. Time to count did not strongly correlate with target number. Performance time for one subject decreased by 60% by altering the frame width. Learning typically, as measured by completion time of individual trials vs. trial number, occurred over 1 to 2 sessions, with r2 values of 0.32, 0.34, and 0.10 and regression line slopes, respectively, of -3.45, -20.0, and -1.62 seconds/trial for each subject. Three subjects correctly counted targets on 25/72, 12/28, and 15/31 sheets containing any number from 1 to 4 targets (p < 0.05, binomial test). One subject also correctly counted targets on 5/8 sheets containing 1 to 8 targets (p < 0.01, binomial test).
Use of a retinal prosthesis can improve subjects’ abilities to identify small, high-contrast targets. Further, conditions for optimal subject performance, such as frame width, vary by subject. The subject pool for this experiment is being expanded to multiple Argus II sites as data collection continues, and a second stage evaluating prosthetic eye-hand coordination will be added.
Clinical Trial: :
This PDF is available to Subscribers Only