Parallel pathways, noise masking and glaucoma detection: behavioral and electrophysiological measures

Purpose: We tested the hypothesis that because of their reduced neural effi ciency, glaucoma patients should have increasingly impaired thresholds as e xternal noise is added to a stimulus. Method: We compared the performance o f 20 normals (mean age = 39 years) with that of 15 patients with early glau coma or at very high risk for glaucoma (mean age 45 years). All patients ha d normal visual acuity. Contrast thresholds were measured on two sets of ta sks: (1) behavioral and (2) sweep visually evoked potentials (VEPs). Two st imuli were used (a) 7.5 Hz reversing gratings of 0.69 cpd, and (b) 5.5 cpd gratings. Noise was binary and contrast varied from 0 to 80%. Psychophysica l thresholds were determined using a staircase which employed a spatial fou r alternative forced choice procedure (4AFC) and converged on 50% correct. Sweep VEP thresholds were determined by extrapolation to zero volts as a fu nction of log contrast. Results: Differences between normal subjects and pa tie nts with early glaucoma were not significant without noise. Both the ab solute size of the difference and its significance increased as noise level increased. For the behavioral thresholds these trends were clearer with th e 5.5 cpd grating, while for the sweep VEPs they were more clear for the 0. 69 cpd grating. Conclusion: The performance deficit of glaucoma patients wh ich may be minimal under normal testing conditions is magnified when extern al noise is added to the stimulus. VEPs and psychophysical thresholds show interesting differences in their sensitivity to this effect. Implications f or the early detection of glaucoma are discussed.