To understand how primary and secondary spherical aberrations affect focusing of the retinal image and the measurement of refractive state in the accommodating eye.Methods
A computational eye model was constructed from published anatomical dimensions of the eye's refractive elements for a range of accommodative states. Two strategies for controlling accommodation were implemented, one in which paraxial rays are always perfectly focused and the other in which paraxial accommodative lag grew larger as target vergence increased. Multiple configurations of the model were achieved by selecting various combinations of pupil size and aberration structure. Refractive state was defined as optimum target vergence for maximizing retinal image quality according to several scalar metrics.Results
When accommodation optimally focuses paraxial rays, retinal image quality is sub-optimal for metrics of image quality sensitive to non-paraxial rays. This loss of image quality can be recovered by optimizing target vergence computationally, which indicates the presence of real accommodative error according to the non-paraxial metric even though the eye is accurately focused paraxially. However, such errors are spurious if non-paraxial refractive state is misinterpreted as paraxial refractive state. Accommodative errors may indicate lag or lead, but in general the slope of the stimulus-response function is less than 1 for non-paraxial measures of image quality. These results depend strongly on pupil size and its variation due to accommodative miosis.Conclusions
spurious accommodative errors can appear when the eye focuses the retinal image optimally according to one metric of image quality (e.g. paraxial) while ocular refractive state is measured by another (e.g. non-paraxial). Spurious errors are small compared to real accommodative lag for small, photopic pupils but can be of the same order of magnitude as real lag for large, mesopic pupils.