Abstract
Purpose :
To compare spectral domain optical coherence tomography (SD-OCT) to adaptive optics OCT (AO-OCT) in the measurement of retinal nerve fiber layer (RNFL) thickness. SD-OCT has been successful in providing detailed measurements, yet the reported axial (8 µm) and lateral (14µm) resolutions prohibit delineating local variations in thickness, such as those due to individual RNFL bundles, and the possibility of identifying non-neural components. The improved lateral (2.4 mm) and axial (4.7 mm) resolution of AO-OCT may provide more detailed information of these local variations and assistance in identifying non-neural tissue.
Methods :
Four control subjects free of ocular disease (ages 24-50) were enrolled and imaged with Spectralis®(Heidelberg Engineering) SD-OCT and the Indiana AO-OCT. With SD-OCT, a 30°x30° central scan was acquired and composed of B-scans spaced at 31 µm. With AO-OCT, 1.5°x1.5° images were acquired at 1°, 3°, 6°, 8° temporal to the fovea in all subjects and additional images at 8° nasal to the fovea in one of the subjects. A-scans and B-scans were spaced at 1 µm. Thickness measurements of the RNFL were extracted from the 3 sections per degree field (totaling 17regions) for SD-OCT and AO-OCT images using ImageJ. Approximately 30 measurements were made per B-scan.
Results :
The AO-OCT images revealed clear delineation of individual RNFL bundles compared with SD-OCT that provided a more uniform appearance of the RNFL bulk thickness. Mean difference across the 17 regions was -3.9 ± 1.8 microns, t = 9, p < 0.0001. The ratio of standard deviations had a mean 1.8 times higher with AO-OCT. In the temporal region where RNFL thins, non-neural components were evident between RNFL bundles due to their decreased reflectance in the AO-OCT images. Average nasal thickness values for SD-OCT and AO-OCT were 52.1mm(SD 5.1), 44.4mm(6.9).
Conclusions :
Similar measurements of the RNFL thickness were found, but AO-OCT measured slightly thinner which may be due to the resolution allowing for better delineation of the bundles. As expected, AO-OCT reveals more localized variations in RNFL thickness. One subject had nasal imaging performed allowing comparison in thicker tissue which still demonstrated the difference in values. The ability to identify some of the non-neural tissue that compose the RNFL may allow for improved understanding of the limits of RNFL thickness as a surrogate to retinal nerve fiber damage secondary to glaucoma.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.