Abstract
Purpose :
Visual performance of intraocular lens (IOL) is often clinically evaluated under binocular condition while majority of optical bench tests are performed monocularly. To translate monocular performance to binocular performance for multifocal IOLs, a gain-control model was developed and cross-validated using existing clinical data.
Methods :
Monocular and binocular defocus curve data were collected in a randomized, parallel-group, subject-masked study that required implantation of the AcrySof IQ ReSTOR +2.5 D multifocal IOL (MIOL) in the dominant eye (DE) and randomization of the fellow non-dominant eye (NDE) to receive either the AcrySof IQ ReSTOR +3.0 D MIOL (contralateral group) or the AcrySof IQ ReSTOR +2.5 D MIOL (bilateral group) in 103 subjects. For both DE and NDE, a stimulatory (k+) and an inhibitory component (k-) was modeled in a gain ratio form (σDE = k-DE/k+DE; σNDE = k-NDE/k+NDE). Another tonic component (c) was included in the gain-control model. Monocular defocus curve was modeled under open-loop conditions while binocular defocus curve was modeled under close-loop conditions. The σDE, σNDE and c are optimized and parameterized by fmincon routine in Matlab and using contralateral group data. The model was further cross-validated using bilateral group data.
Results :
The inhibitory/stimulatory gain ratio was 0.686 for the DE and 0.474 for the NDE. R2, as estimates of goodness-of-fit, were 0.853 and 0.924 respectively for contralateral (n=50) and bilateral (n=53) groups. Averaged prediction errors at all defocuses were within 0.06 logMAR for the contralateral group. For the bilateral group, the prediction errors were within 0.06 logMAR from +2.0D to -3.0D and became larger from -3.5D to -5.0D (up to 0.12 logMAR).
Conclusions :
The crosslink gain-control model can provide a generalized binocular model that accounts for the interaction between the eyes including binocular summation and rivalry. The model can be used to predict binocular visual performance such as visual acuity and defocus curve.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.