Imaging the foveal cones in vivo through ocular speckle interferometry: theory and numerical simulations

Abstract

New high-resolution imaging techniques have recently been developed for resolving the foveal cone mosaic in the living human eye. We present a theoretical analysis and numerical predictions of the performance and limitations of ocular speckle interferometry and related methods. Images were computer generated, simulating short-exposure recordings of a model of the cone mosaic imaged through the optics of the eye. Cases of both coherent and incoherent illumination are analyzed. Theoretical predictions are in good agreement with published experimental results. Our analysis and computer simulations validate these methods and successfully explain different observed effects. The simulation predicts a slight underestimation of the experimentally measured cone mosaic characteristic spatial frequency and how this effect can be compensated by deconvolution with a speckle transfer function. Other experimental observations, such as the improvement of the results after a preselection of the best images, the effect of both the defocus and changes in fixation, and the influence of the size of the illuminating spot, have been reproduced.