Abstract
Purpose :
Evaluate the robustness of the Direct Subjective Refraction (DSR) method against the stimulus design and the parameters of the test.
Methods :
The DSR is a new subjective refraction method that projects a tunable lens onto the eye’s pupil to change the focus periodically at 15Hz with 0.25D of amplitude. The subject looks through the optical system to a stimulus made of two monochromatic components and perceives flicker and chromatic effects. Their task is to minimize those perceptual cues. The DSR method was performed for different temporal frequencies (10, 15, 20Hz), defocus amplitudes (0.125, 0.25, 0.5D), and stimulus designs (number of monochromatic circles or bars; the color of the monochromatic components: Blue&Red -BR-, Red&Green -RG-, Green&Blue -GB-; and presence of bichromatic components). Stimuli was projected (DLP projector) on a white screen. Five subjects (27±3 yo) performed the DSR task in one eye with their current prescription, repeating 4 times each condition to provide the refraction, measurement time, and repeatability.
Results :
The time per repetition was 14±6s, averaged across repetitions, conditions, and subjects, and was stable across them. The mean repeatability was 0.17±0.16D, and the lowest 0.06D (GB, circles and ring). The spherical equivalent (M) obtained with circles was -0.7±0.36, -0.31±0.37, and -0.69±0.35D for BR, RG, and GB. Results with BR and GB stimuli were not statistically different (p>.05) but with RG were significantly higher (0.4D, p<.05). Bars allowed to measure astigmatism besides M. Average astigmatism was subclinical 0.18±0.16D, without differences among color components (p>.05). The M from bars was similar to that from circles (p<.05). Again, RG was statistically different to BR and RG. The longitudinal chromatic aberration (LCA) of the eye can explain the difference obtained with longer wavelengths (RG vs BR) but explaining the similarity of BR and GB requires additional perceptual effects or accommodation. Temporal frequency and defocus amplitude did not affect the result (p<.05) but subjects preferred 15Hz and 0.25D.
Conclusions :
The DSR results are stable against the parameters of the parameters of the test and the design of the stimulus. DSR is affected (~0.4D) by the chromatic components of the stimulus, with a fixed -and predictable- offset in good agreement with the LCA.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.