Abstract
Purpose :
Gliosis-like hyper-reflective lesions near the inner limiting membrane (ILM) are found in patients with glaucoma (Graf et al., J. Glaucoma 1993; 2:257-9), and in older healthy eyes (Gast TJ, et al. IOVS 2013;54:ARVO E-Abstract 1507). Our goal was to reduce the impact of putative glial refllectance on en face images extracted from optical coherence tomography (OCT) volume scans as slabs averaged across a range of distances from the ILM (Hood et al., IOVS 2015; 56:6208-16).
Methods :
29 patients (20 with glaucoma and 9 glaucoma suspects) and 18 age-similar controls were prospectively imaged on the Spectralis with 4 high-density OCT scans (vertical scans with 30 micron spacing) that covered most of the retina tested with 24-2 perimetry. Data from controls were used to assess lesion depth, and choose custom slabs. Slab analysis was performed for the patient data with the 0-52 micron slab proposed by Hood et al 2015, and with slabs customized to reduce the impact of glial reflectance: 16-28 microns below the ILM for the temporal raphe, 24-36 microns for the parafovea, and 32-52 microns for the arcades and disc. Reflectance for each slab was computed by averaging across depths in two ways: for absolute reflectance values returned from the manufacturer’s transform, and for relative reflectance computed as a fraction of reflectance at Bruch’s membrane.
Results :
19 patients and 13 controls had putative glial lesions near the ILM. For controls the lesions were 16-24 microns deep, while maximum RNFL thickness was 28-36 microns for the raphe, 48-72 microns for the parafovea and 160-344 microns for the arcades. For 12 patients the lesions caused substantial reflectance artifacts with the Hood slab (0-52 microns). For these patients our custom slabs removed artifacts at the disc and arcade for all 7 patients with artifacts there for the 0-52 micron slab, in the parafoveal region for 5 of 10 patients, and in the temporal raphe for 0 of 7 patients. Of the remaining 12 cases, our custom slabs yielded milder artifacts for 5 cases and similar artifacts for 7 cases.
Conclusions :
Lesions due to putative gliosis are common in older eyes, and slab parameters for en face RNFL images can be chosen to reduce the impact artifacts from these lesions in the arcades and disc, but not in thinner regions of the RNFL.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.