Abstract
Purpose: :
Accurate knowledge of spatial summation among foveal cones is essential for proper interpretation of recent adaptive optics (AO) psychophysical results. Unfortunately, due to optical limitations, prior measurements have only provided an upper bound. AO offers the potential to more precisely determine the spatial extent of pooling among foveal cones. Towards this end, we investigate foveal cone pooling at absolute threshold by measuring dark-adapted thresholds with and without AO correction of the eye’s aberrations.
Methods: :
Dark-adapted foveal detection thresholds were measured for two subjects when viewing a brief (6 msec), monochromatic (580 nm, 20 nm bandwidth) point source, displayed with conventional optics (2mm pupil, best refraction) and with AO correction of the eye’s aberrations (6mm pupil). Guessing corrected thresholds were interpolated at 50% seen for multiple response criteria (rating scale method) and compared between conditions. Results were compared with simulations for different pooling and detection scenarios.
Results: :
Stimulus full width at half height with AO was 0.40' for both subjects, compared to 1.1' with conventional optics. Despite a 7.5-fold decrease in stimulus area, AO corrected thresholds increased 80 ± 11% and 56 ± 10% above conventional thresholds, from 415 to 750 and 541 to 846 corneal photons, for subjects 1 and 2 respectively. AO corrected thresholds were also consistent with increased detection uncertainty.
Conclusions: :
At first pass, an increase in thresholds with a smaller (AO corrected) stimulus appears inconsistent with simple models with spatial pooling less than, or equal to, previously established upper bounds. However, spatial inhomogeneity in the retinal mosaic as well as inhibition may elevate threshold for the smaller stimulus, even in the presence of reduced spatial pooling. Furthermore, increased detection uncertainty for the AO corrected stimulus, which may reflect increased uncertainty in spatial localization, is expected to further elevate threshold. Detection models are used to further clarify the impact of detection uncertainty and the mosaic on thresholds.
Keywords: photoreceptors: visual performance • detection