Purpose: To examine the potential clinical utility of equiluminant chromatic stimuli for assessing glaucomatous damage. Pan et al. (2006) found that equiluminant red-green chromatic stimuli could have good ability to detect defects as well as low test-retest variability, but clinical utility was limited due to the small dynamic ranges for their stimuli. The current study increased the dynamic range by using larger stimuli and including tritan stimuli.
Methods: Luminance, red-green (R-G), and tritan stimuli were created by modulating a large square (3 degrees per side) from an equal energy white (20 cd/m2) along three cardinal directions in color space. Contrast sensitivity was measured at four locations with an eccentricity of 12 deg. along the 45, 135, 225, and 315 degree meridia. Twenty-five patients with glaucoma and twenty-six control subjects free of eye disease were tested monocularly at two separate sessions within a two-week time period. Sensitivities were reported in decibel (dB) units, where 1 dB=-1 (log contrast threshold) x 10.
Results: The dynamic ranges were 11 and 13 dB for the tritan and red-green stimuli. Test-retest variability was dependent on depth of defect for the two chromatic stimuli (r>0.2, p<0.25) but not for the achromatic stimuli (r=0.01, p=0.46). Matched t-tests found that, on average, defect depths were similar for the red-green and luminance stimuli (t=0.5, p=0.30), and were slightly deeper for the luminance stimulus than for the tritan stimulus (t=4.2, p<0.0001). The relationship between defect depths for the luminance and tritan stimuli was dependent on mean defect (r=-0.42, p<0.0001).
Discussion: The effort to increase the dynamic range for the chromatic stimuli by increasing the size of the stimulus and using tritan modulation were successful. However, this came at the cost of increased test-retest variability and decreased ability to detect glaucomatous defects.
Conclusion: Equiluminant chromatic stimuli in CRT-based tests may not be clinically useful as perimetric stimuli, since increased dynamic range comes at the expense of increased test-retest variability and decreased ability to detect visual loss. Those findings further support the works of Hart (1988) and Sample et al. (2006).