Abstract
Purpose:
Real-time analysis of distortion in Scanning Laser Ophthalmoscopy (SLO) videos can accurately record eye motion at high frequency. This facilitates repeated delivery of stimuli to targeted locations on the moving retina. The technology was developed for an Adaptive Optics SLO (AOSLO) and has more recently been implemented in a simpler system without AO, the Tracking SLO (TSLO). The purpose of this experiment was to investigate the role of natural and manipulated retinal image motion, or lack thereof, on stimulus contrast thresholds and show how the measurements compare with AOSLO results.
Methods:
The threshold for detection of a diagonal grating stimulus that increased in contrast over a 6 second trial was measured using TSLO with a 2 AFC judgment of orientation. Three subjects were tested under 7 motion conditions at the fovea and 4 subjects were tested at 2° and 5° eccentricity. Three cycles of a 6cpd grating were encoded directly into a 4° 840nm imaging square. The change in contrast at the fovea and 2° was 0.1-0.7. At 5° this was increased to 0.3-0.9 or 0.4-1 to make the task easier. Stimulus motion conditions included natural, stabilized, and amplified/minified motion in directions consistent with and opposite of natural eye motion.
Results:
The highest contrast thresholds were measured under the stabilized stimulus condition for all locations tested. Conditions in which the retinal motion was in a direction consistent with natural motion resulted in slightly lower thresholds. These trends are consistent with previously reported results obtained using AOSLO for smaller targets in a smaller field, demonstrating it is possible to study the role of image motion using a TSLO system.<br /> <br /> Advantages of TSLO include a larger field size, which enables the use of larger stimuli and tracking of larger eye motions. TSLO is also far less costly, more compact, and can be built and operated with less training. The disadvantage is the lack of AO, resulting in lower image quality that varies with aberrations in individual eyes, compromising tracking accuracy and precluding fine optical control of the retinal stimulus.
Conclusions:
TSLO is as good or better than AOSLO for running psychophysical experiments exploring the role of retinal image motion. The advantages of TSLO make it an economical solution that could facilitate wider-spread studies where detailed real-time cellular-level retinal detail is not crucial.