Prediction of Accommodative Optical Response in Young and Pre-Presbiopic Human Eyes Using Ultrasound Biomicroscopy

Purpose: Clinical accommodation testing involves measuring either the accommodative optical response (AOR) or the accommodative biometric changes in the ocular anterior segment. Currently, it is not possible to measure both with a single instrument. Measuring the AOR and the accommodative biometric changes are important for evaluating accommodation restoration concepts. The specific goals of this research are: 1) to perform automated, objective measurements of accommodative biometric changes from ultrasound biomicroscopy (UBM) images in young phakic eyes; 2) to measure the static AOR using a Grand Seiko (GS) autorefractor and infra-red photorefraction (PR) in young eyes and to predict the AOR from UBM measured biometric changes; 3) to measure the AOR using GS and PR and the biometric changes using UBM in pre-presbyopes to predict the AOR; 4) to calculate and correct the spatial and optical distortion in Visante optical coherence tomography corneal images; 5) to construct accommodative schematic eye models for individual eyes for each of the young and pre-presbyopic subjects and calculate refraction and AOR from the schematic eye models.

Methods: Experiments were 1) Accommodative anterior segment biometric changes were measured in response to 0 D to 6 D accommodative stimuli in 1 D steps in 26 young human subjects using a 35 MHz UBM and an A-scan ultrasound. 2) Static AOR to the same stimulus demand were measured with GS and PR in the same group of young subjects. AOR was predicted from UBM measured biometry parameters using linear regression, 95% confidence intervals and 95% prediction intervals. 3) Static AOR to 0 D to maximal stimulus demand in at least 0.25 D steps was measured with GS and PR in 25 pre-presbyopic human subjects. Accommodative anterior segment biometric changes were measured using UBM and A-scan ultrasound. AOR was predicted from UBM measured biometry parameters as described in experiment 2. 4) Five contact lenses of known front and back surface radii of curvature and central thicknesses were imaged using the Visante to calculate spatial and optical distortion corrections which were then applied to corneal images captured from the young and pre-presbyopic subjects. 5) Ocular biometry parameters (Visante, A-scan and UBM) from Experiments 1, 3 and 4 in young and pre-presbyopic subjects were used to construct paraxial schematic eye models for each individual subject for each accommodative stimulus demand.

Results: 1) Standard deviations of UBM measured parameters were smaller than A-scan measures. 2) Mean prediction errors of AOR using linear regression in young subjects for various biometry parameters ranged from 0.56 D to 0.91 D. 3) Mean prediction errors of AOR using linear regression in pre-presbyopic eyes ranged from 0.41 D to 0.62 D. 4) Root mean square (RMS) error of the power of the contact lens surfaces after distortion correction was 0.18 D for the front and 0.11 D for the back surfaces, respectively. 5) Mean ± SD of prediction errors of AOR from individual schematic eyes for the young and pre-presbyopic subjects were 0.50 ± 0.39 D and 0.50 ± 0.37 D, respectively.

Conclusions: The results show: 1) the utility of automated image analysis to get accurate, rapid and objective measurements of anterior segment biometry from UBM images; 2) how spatial distortion in UBM images can be corrected to get accurate measurements and the ability of each UBM measured biometry parameter to predict the AOR; 3) how UBM, despite having low axial resolution, can predict AOR in pre-presbyopic eyes with low accommodative amplitudes; 4) how spatial and optical distortions in Visante images can be corrected to get accurate corneal biometry that can be used for schematic eye modeling; 5) how individual schematic eye predictions of the AOR are better than predictions using LR from individual UBM measured biometry parameters.