Purchase this article with an account.
Pete R Jones, Sarah Kalwarowsky, John Wattam-Bell, Marko Nardini; An automated test of infant visual acuity using remote eye-tracking. Invest. Ophthalmol. Vis. Sci. 2014;55(13):2737. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To compare binocular visual acuity (VA), estimated by: (i) experimenter judgments of preferential looking (the current gold standard); (ii) a new automated computer-based system.
VA was estimated in 27 infants (3 - 12 months), using both Keeler Acuity Cards (Keeler Ltd., UK) and an automated procedure based on remote eye-tracking. In the automated test, the infant was sat on their parent’s lap, and a black-and-white grating was presented on an LCD screen. The grating was located in a random position, eight degrees of visual angle from the current point of fixation. The infant was scored as having looked at the grating using an automated algorithm, based on data from an eye-tracker mounted below the monitor (Tobii X120; Tobii Technology AB, Sweden). After each trial, the spatial frequency of the grating was adapted up/down, in order to determine the infant’s detection threshold.
All infants successfully completed each test at least once during a single session. There was good agreement between the mean VA estimates from the acuity cards and the eyetracking procedure [r = 0.6, p < 0.01]. There was no significant difference in test duration [t = 1.3, p = 0.20], with each test taking approximately 2 - 3 minutes. However, the mean test-retest discrepancy in VA was greater for the automated procedure (1.6 cycles/deg) than with acuity cards (0.8 cycles/deg). Possible reasons for this will be discussed.
A fully automated test can provide a viable measure of infant acuity, and will benefit from ongoing additional development. Use of electronic displays and eye tracking provide flexibility in the design and choice of visual stimuli. In the future these methods have the potential to measure specific aspects of vision, relevant to particular clinical groups, using stimuli that would not be otherwise feasible. Benefits (ease of use, cost, extensibility) and limitations (portability, initial outlay) of an automated system will be detailed.
This PDF is available to Subscribers Only