Different schematic eyes and their accuracy to the in vivo eye: a quantitative comparison study

Abstract

Current ophthalmic technology allows the manipulation of eye components, such as anterior cornea and lens, of the human eye with a considerable precision and customization. This technology opens up the possibility of exploiting some characteristics of the eye in order to improve the methods of correcting optical aberrations. Moreover, product development and research for the eye-care professional has reached very high standards, since there is nowadays software available to design and simulate practically any mechanical or optical characteristic of the product, even before it is thrown into production line. Although quite similar in the general form, different human eye models simulate the image formation by considering different property combinations in the constitutive elements of the eye structure (such as refraction index and surface curvatures), producing retinal images that resemble very closely those of the biological eye. Using optical design software, we have implemented a simulation of 5 well-known schematic eyes available in the literature. These models were the Helmholtz-Laurance, Gullstrand, Emsley, Greivenkamp and Liou & Brennan. The optical performance of these different models was compared using different quantitative optical quality parameters. The model of Liou and Brennan, contains features of the biological eye that were not considered in previous models, as the distribution of a gradient refraction index and a decentered pupil. Furthermore, it has great reliability since it takes into account the mean value of empirical measurements of the in vivo eye in order to define size and parameters such as anterior and posterior curvature of cornea, lens, axial length, etc. Comparisons between the MTF (Modulation Transfer Function), spot diagrams and ray fan showed the difference in image quality between eye models, and the Strehl Ratio was also used as a parameter of comparison. A careful comparison between the different models showed that the first four schematic eyes have better optical quality than what is expected for the general and healthy emmetropic in vivo eye. Liou and Brennan schematic eye is the one that most closely resembles the in vivo biological eye. Therefore, in applications, such as research or product development for customized vision correction, which must consider optical properties intrinsic to the biological eye, we recommend this latter model; for applications that do not require refraction-limited performance, most of the other models should be a good approximation.

Variability in Human Cone Topography Assessed by Adaptive Optics Scanning Laser Ophthalmoscopy

“…31 This distance was calculated with the Liou and Brennan eye model, 41 which closely resembles in vivo biological eye. 42 Model parameters, including anterior radius of curvature of the cornea (r1), anterior chamber depth (ACD), and axial length, were measured from each subject’s eye with the IOLMaster. Corneal thickness was subtracted from the measured ACD to obtain the distance from the posterior cornea to the anterior lens.…”

Personalized Pseudophakic Model for Refractive Assessment

“…It is a finite, non-paraxial model, being the most realistic concerning the average of aberrations of the physiologic eye [14]; [15]. As defined by this model, pupil decentration was set at 0.5 mm from the optical axis with a 5° angle between the visual and optical axis.…”

Influence of transverse mode on retinal spot size and retinal injury effect: A theoretical analysis on 532-nm laser

“…21,30 In the model, the following parameters are involved: the curvature of anterior and posterior surfaces of the cornea and lens, the thickness and the refractive index of the cornea, aqueous, lens and vitreous. All the interfaces in the eye model are ideal spherical surfaces.…”