A computer model for predicting image quality after photorefractive keratectomy

Accurately predicting visual performance remains a concern in refractive surgery. The effects of the eye's optics on retinal image quality were investigated using computer ray tracing to model the human eye after photorefractive keratectomy (PRK).

Ray-tracing analysis was used with an anatomically realistic model of the human eye including aspheric surfaces and crystalline lens gradient index distributions. The contribution of corneal curvature to refractive error was investigated using data of axial length, corneal power, anterior chamber depth, and lens power from 318 eyes from the literature. The computer interface was specifically designed for use with PRK and provides graphical plots of the remodeled eye, ray paths and retinal image formation.

Modeling the optical contribution of corneal curvature resulted in an improvement in predicted refractive state of the eye as a function of axial length expressed as the R2 value of the regression analysis from 0.88 to 0.96. Subsequently, analyses were conducted for single and multizone treatment areas of differing diameter and with varying pupil size. Retinal image quality following PRK for the human cornea was found to be affected by not only the corneal parameters of anterior curvature and thickness, but also by axial length, pupil size, and anterior chamber depth.

The inclusion of multiple interdependent optical parameters showed differences from conventional methods in predicting refractive outcome following PRK and revealed factors affecting image quality may account for some imperfections in visual performance based on simpler optical modeling.