Purpose
We investigate the correction strategy in an adaptive optics system for retina imaging.<br /> To reduce the cost and complexity of current adaptive optics systems, it is necessary to assess which specifications can be loosened without performance losses. To do so, one has first to study the perturbations to be corrected, and then derive the most suited correction scheme. We focus here on the dynamic aspect. In particular, as ocular movements are among the main contributors to ocular dynamic aberrations, we explore the potential of pupil stabilization in AO-assisted systems dedicated to retina imaging.
Methods
We examined dynamic aberrations simultaneously with pupil movements on a 50-eye non pathological population with a high resolution custom-built biometer. The biometer consists of a Shack-Hartmann wavefront sensor synchronized with a pupil camera and is running at 236Hz.<br /> In parallel, we set up an adaptive optics retina imaging system featuring an 88 actuator deformable mirror with pupil stabilization and running at more than 80Hz, in order to quantify the influence of pupil stabilization as well as the impact of the frame rate on the AO correction performances.
Results
We highlighted a correlation between micro-saccades and the level of aberrations. Yet, we showed that correcting dynamic aberrations from pupil shifts does not fully correct for the aberrations implicated with pupil shifts: other sources of ocular movements induced aberrations exist apart from horizontal and vertical eye rotations (for instance, lens wobbling).<br /> Undergoing tests on our AO-assisted retina imaging system will enable to conclude on the efficiency of the pupil stabilization to compensate for dynamic aberrations for various deformable mirror speeds.
Conclusions
We demonstrated on several subjects that a significant part of the aberration dynamics cannot be explained by the combination of static aberrations and pupil motion. We will discuss at ARVO the impact of frame rate and pupil stabilization on AO performance.