Abstract
Purpose :
Convolved images appear always more degraded than the same image blurred with optical defocus. We have explored the role played by accommodation (i.e., microfluctuations of accommodation (MF) and residual accommodation) in this discrepancy.
Methods :
Seven perfectly corrected (best focus) young subjects compared (time-based comparison) images degraded by simulated dioptric blur with real defocused images (from 0 to +1 D in a 0.125 D step) seen through corresponding optical lenses. The convolved images (convolution between an original image and a point spread fonction filter) were seen at 4.18-m through a 3-mm pupil size to limit the impact of the observer’s aberrations whereas real blurred images were viewed through a 5-mm pupil size. Original Images were either three high-contrast 0.4 LogMAR letters or a digitally created "Face" image. The comparison were performed in monochromatic and polychromatic light. For both conditions, subject aberrations may or may not have been taken into account in the convolution filter. Accommodation was dynamically recorded (WAM-5500) before and after the comparison task. We did not use cycloplegia.
Results :
A repeated measures ANOVA showed significant differences considering or not aberrations (p=0.006 Lens*With or Without Aberrations) but no effect of the target. Without considering accommodation, the difference between simulated and real blurred images is -0.14 ± 0.16 D and -0.06 ± 0.19 D (mean ± SD) respectively in polychromatic and monochromatic conditions. Simulated images appeared more degraded. When taking into account the difference of accommodation between the no-defocus condition (best focus) and the defocus lens one, the difference is 0.06 ± 0.13 D and -0.05 ± 0.28 D respectively in polychromatic and monochromatic conditions. The Bland-Altmann analysis confirms the benefit of considering the accommodation. MF were quite comparable between subjects and conditions (0.10 D in average) and can not explain the residual difference between real and simulated blurs.
Conclusions :
A low level of residual accommodation had to be considered when comparing real and simulated defocus blur. In doing so, the residual difference between the two is negligible. We have confirmed the importance of taking into account the subject's aberrations in the simulation.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.