Abstract
Purpose :
Numerous gradient index (GRIN) lens models have emerged in recent decades, but the optimal fit for the human lens's dynamic variations and population diversity remains elusive. This study compares the optical performances of four crystalline lens models: homogeneous, GRIN, GRINCU singlet, and GRINCU doublet. Due to the unavailability of comprehensive public biometric datasets that include lens biometry, we implement these models using SyntEyes, which incorporates realistic variations in the human eye to compensate for fixed dimensions.
Methods :
We utilized biometric data from 150 SyntEyes samples, including corneal and lenticular curvature radii, surface thickness, lens tilt and shift, and lens equivalent refractive index. We used custom ray tracing software in Matlab and introduced three new optical surfaces: a GRIN lens, a GRINCU singlet, and a GRINCU doublet. The GRINCU crystalline lens model incorporates a curvature gradient parameter (G) of the iso-indicial surfaces. Finite ray tracing was performed on the 150 SyntEyes samples using the four lens models, with varying G values (-2, -1.5, -1, 1) applied specifically to GRINCU singlet and doublet configurations.
Results :
We computed the average cardinal points and Zernike aberration coefficients (Figure 1) for each SyntEyes sample, as well as the cardinal points specifically for each lens model. As the curvature gradient (G) becomes increasingly negative, the average power of the GRINCU lenses decreases—from 22.57 ± 0.33 D for G=1 to 22.41 ± 0.32 D for G=-2—, resulting in an average forward shift of their principal planes by roughly 180 ± 33 microns and a slight modification in the distance between them. This displacement of principal planes is also observable in the entire eye.
Conclusions :
A negative curvature gradient (G) in GRINCU lenses decreases lens power, causing a forward shift of both the lens and the eye’s principal planes. Lower G values lead to reduced defocus and primary spherical aberration.
This highlights the complexity and variability of human lens dynamics and presents a step toward understanding the optimal fit for population diversity and dynamic variations.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.