Purchase this article with an account.
Michael Doyle, Gary C Lee, Niranchana Manivannan, Keith O'Hara, Mary K Durbin, Marie-Noëlle Delyfer, Jean-Francois Korobelnik; Widefield Fundus measurements on Argus II implant. Invest. Ophthalmol. Vis. Sci. 2020;61(7):5274.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The Argus II retinal implant, with its grid of electrodes at known spacing, presents the unusual opportunity of measuring absolute lateral distances at the retina. Usually, the optical magnification of each subject’s eye is replaced by a conventionally assumed magnification, to allow repeatable but not necessarily accurate measurements on the retina. This retrospective study reports inferred magnifications based on measurements taken on a commercial fundus camera in two eyes with the Argus II retina implant.
Widefield, true-color fundus images acquired using CLARUS™ 500 (ZEISS, Dublin, CA) on two eyes of two subjects with Argus II implants were retrospectively analyzed. Centroids of visible electrodes were manually annotated in MATLAB (MathWorks, Natick, MA). The CLARUS measurement algorithm, which is based on the angles between rays mapped from individual pixels multiplied by the optical magnification established in ISO 10940 (17mm/rad), was used to calculate the distances between each pair of electrodes and the inferred magnifications were compared to the results expected based on the nominal pitch.
Distances between 496 pairs of visible electrodes in Image 1 and 1,035 in Image 2 were calculated. Figure 1 shows both images with insets illustrating where the electrodes' centroids were assigned. The image quality for Image 1 was especially poor. Figure 2 shows the inferred magnification of the individual eye vs. feature size. In Image 1, inferred magnification ranged from 18.25 to 21.27 mm/rad, median 19.46mm/rad. In image 2, inferred magnification ranged from 15.90 to 19.31 mm/rad, median 17.55 mm/rad. The scatter in inferred magnification decreased relative to increasing distance between electrodes.
In this study, the differences in magnification range between the two images suggest measurement accuracy varies with each eye, likely due to differences between the true magnification and the conventional magnification model; it's also possible the implant has deformed during the implantation process. The measurements also assume reasonable image quality as well as proper fixation, which may be difficult for patients requiring Argus implants. Since commercial fundus cameras, such as CLARUS, typically base measurements on an average optical model (i.e., not customized for an individual eye), measurement repeatability1 may be more clinically relevant than measurement accuracy.1Leahy et al, 2017. IOVS, 58(5453)
This is a 2020 ARVO Annual Meeting abstract.
This PDF is available to Subscribers Only