Abstract
Purpose: :
To develop and test the application of an adaptive optics scanning laser ophthalmoscope (AOSLO) with eye tracking for high-resolution microperimetric testing of normal subjects.
Methods: :
An AOSLO was used to conduct simultaneous retinal imaging and visual function testing in five normal subjects. Imaging was conducted at four retinal locations (fovea, 0.75º, 2.25º, and 3.75º) along the temporal horizontal meridian over the course of two measurement sessions. Retinal videos were stabilized in real time using a high-speed eye tracking algorithm, enabling the targeted delivery of aberration-corrected Goldmann I-sized stimuli (diameter = 6.5 arcminutes; wavelength = 680nm) to prespecified retinal locations. Increment thresholds were assessed using an adaptive Bayesian staircase procedure (QUEST) and plotted as a function of eccentricity from the subject's preferred retinal locus of fixation (PRLF). Montages of the photoreceptor mosaic were generated for each subject from the stabilized videos and the precise stimulus location for each trial was plotted onto the image. Predetermined exclusion criteria were applied to identify off-target stimuli. These trials were subsequently removed from offline analyses and the associated threshold measures were recomputed. Finally, as further proof of concept, a 4.2 arcminute stimulus was used to probe the physiologic microscotomata associated with the parafoveal vasculature.
Results: :
As expected, retinal sensitivity decreases with increasing retinal eccentricity. The relationship between the square root of the increment threshold and retinal eccentricity was linear (R² = 0.58). In general, the eye tracking and targeted stimulus delivery algorithms performed well. Across all test locations and subjects, the vertical and horizontal stimulus delivery errors averaged 0.81 and 0.89 arcminutes (~4 microns), respectively. Based on our exclusion criteria, the stimulus was successfully delivered to its targeted location on 90.1% of all trials. Increment thresholds measured over parafoveal blood vessels were significantly higher (p<0.05; one-tailed t-test) than those obtained from directly adjacent retinal areas.
Conclusions: :
AOSLO-based microperimetry has the potential to conduct tests of visual sensitivity on a microscopic scale with an unprecedented degree of retinotopic precision. This approach may prove useful to researchers interested in establishing the functional correlates of photoreceptor mosaic structure in both normal and diseased retinas.
Keywords: perimetry • photoreceptors: visual performance