Purchase this article with an account.
E. Dalimier, C. Dainty; Model-Observer Performance for Adaptive Optics Corrected Functional Vision. Invest. Ophthalmol. Vis. Sci. 2008;49(13):996. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Adaptive Optics (AO) techniques have recently been used to highlight the impact of higher-order (HO) ocular aberrations on vision. It has also been shown experimentally that neural contrast sensitivity can moderate the optical blur at low light level (IOVS 48, E-Abstract 1512, 2007). The purpose of this study was to derive a multi-task ideal observer model which accounts for individual optical filtering and light-dependent neural filtering, to understand better their joint effect on visual performance.
The work was based on a classical multi-class ideal observer model. Customised optical filtering was calculated from measured ocular wavefront errors and light-dependent neural filtering was constructed from weighted visual channels. A spatial frequency analysis was performed on the visual stimulus to qualitatively discuss the joint effect of neural and optical contrast sensitivity on the observer performance. The numerical performance of the model-observer was computed with a multi-class separability measure and related to the experimental results recently obtained for a multi-class functional sensitivity visual test. The results were compared in terms of the AO visual performance ratio (ratio of the performance with AO correction of HO aberrations to that without correction), for seven different subjects, and a set of pupil sizes and light levels.
The model numerical results showed good agreement with the experimental results for the set of seven subjects, and for the different pupil sizes. The numerical results, as well as the spatial frequency characteristics analysis, also supported the trend of the AO visual performance ratio to decrease as the light level is decreased.
Although some more individual data (concerning the neural contrast sensitivity) is required for a full quantitative validation, the model showed good agreement with previously obtained experimental results. In particular, it supports the analysis that at low retinal illuminance levels, neural factors limit the effect that increased higher-order aberrations can have on functional visual performance.
This PDF is available to Subscribers Only