Abstract
Purpose :
Color vision deficiency(CVD) tests are a critical part of a vision exam. Conventional tests are time-consuming, require trained clinicians to administer and interpret results, and may only use pass/fail criteria to assess vision loss. We used AIM(Angular Indication Measurement) Color Detection and Discrimination to assess color vision, validated AIM Color in 8 color normal(CN) and 1CVD observers, and compared against Rayleigh matches, HRR, and FM100 hue test.
Methods :
Each task presented charts in different colors 3 times(later charts were responsive-adaptive to previous charts). Each chart contained 4×4 5° cells with a 2° chromatic stimulus, randomized in orientation and signal intensity, centered in each cell and embedded in 20% contrast 14 Hz dynamic luminance noise. Detection stimuli were Cs in L,M,S-cone isolating directions. Discrimination stimuli were circular patches in which each half contained a different color. 4 primary(S+, S-, red & green) and 4 intermediate equiluminant color directions were chosen as test colors. The two colors in the patch were selected around each test color. The location of each C’s gap or the central edge orientation was indicated by observers by clicking the corresponding location on the surrounding ring. AIM thresholds, deterioration of vision(slope), noise(min. error), and bias(target orientation vs. observer error) were calculated.
Results :
All CNs passed the three standard tests. The CVD observer was classified as protanomalous by the anomaloscope, mild red-green deficient by HRR, and protan deficient by FM100. Results show that the protanomalous observer had detection and discrimination thresholds that deviate from those of the CNs, and significant differences(p<.001) were found for LM detection conditions and S+, S-, green, and S-/red discrimination conditions. Mean test times were 12.5min(n=8), 46.8s(n=8), 13.0min(n=9), 30s & 36s(n=9) for Rayleigh, HRR, FM100, 1 AIM detection & 1 AIM discrimination chart, respectively. Noise and bias were not significantly different for both tests in CN but different when compared to the CVD observer. Slope showed an effect among discrimination colors.
Conclusions :
AIM Color enables a rapid, self-administered, and personalized interrogation through response-adaptive approach and provides additional diagnostics. Preliminary results suggest that threshold elevations were observed for the CVD observer in both AIM detection and discrimination.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.