Evaluation of novel tablet-based color vision tests

Poster Presentation 33.340: Sunday, May 19, 2024, 8:30 am – 12:30 pm, Banyan Breezeway
Session: Color, Light and Materials: Neural mechanisms, models, disorders

Christabel Arthur1 (), Jingyi He2, Peter J. Bex3, Jan Skerswetat3, Michael A. Crognale1; 1University of Nevada, Reno, 2University of California, Berkeley, 3Northeastern University, Boston

Color vision testing has shown significant advancement throughout the years, with a recent shift to computerized and self-administered testing, including the application of adaptive psychophysical testing and automated scoring. Here, we report a preliminary evaluation of the AIM (Angular Indication Measurement) and FInD (Foraging Interactive D Prime) color vision tests, part of a tablet-based diagnostic suite offered by PerZeption (Boston, MA). We also ran a full battery of standardized color tests on the subjects and specifically compared results from the AIM and FInD with those from the “trivector” test of the Cambridge Colour Test (CCT), and the Mollon-Reffin test to determine their relative effectiveness in detection and classification of color vision deficiencies (CVDs). These tests were chosen for comparison because they each quantify discrimination along the individual cone axes. Initial diagnosis was determined with an anomaloscope. There were two AIM paradigms, a detection task in which participants identified the gap of Landolt C stimuli and a discrimination task in which the orientation of bi-partite stimuli was identified. For the FInD detection paradigm, participants chose those boxes out of an array of boxes in which they detected a color blob. For the FInD discrimination task, they chose boxes out of the array in which two blobs of color were discriminable in color. AIM and FInD color threshold estimates were recorded for observers with normal color vision and with CVD. As expected, congenital red-green color deficient subjects had significantly higher thresholds on L and M cone and L-M (AIM and FInD) tests than did the control group. Results of both of these tests correlated reasonably well with those from the CCT and the Mollon-Reffin tests. Our results suggest that AIM and FInD provide reliable CVD diagnostics.