Estimation of the central 10-degree visual field using en-face images obtained by optical coherence tomography

Abstract

To estimate the central 10-degree visual field of glaucoma patients using en-face images obtained by optical coherence tomography (OCT), and to examine its usefulness. Patients and methods Thirty-eight eyes of 38 patients with primary open angle glaucoma were examined. En-face images were obtained by swept-source OCT (SS-OCT). Nerve fiber bundles (NFBs) on enface images at points corresponding to Humphrey Field Analyzer (HFA) 10-2 locations were identified with retinal ganglion cell displacement. Estimated visual fields were created based on the presence/absence of NFBs and compared to actual HFA10-2 data. κ coefficients were calculated between probability plots of visual fields and NFBs in en-face images. Results Actual HFA10-2 data and estimated visual fields based on en-face images were well matched: when the test points of <5%, <2%, and <1% of the probability plot in total deviation (TD) and pattern deviation (PD) of HFA were defined as points with visual field defects, the κ coefficients were 0.58, 0.64, and 0.66 in TD, respectively, and 0.68, 0.69, and 0.67 in PD. In eyes with spherical equivalent � −6 diopters, κ coefficients for <5%, <2%, and <1% were 0.58, 0.62, and 0.63 in TD and 0.66, 0.67, and 0.65 in PD, whereas for the myopic group with spherical equivalent < −6 diopters, the values were 0.58, 0.69, and 0.71 in TD and 0.72, 0.71, and 0.71 in PD, respectively. There was no statistically significant difference in κ coefficients between highly myopic eyes and eyes that were not highly myopic. Conclusions NFB defects in en-face images were correlated with HFA10-2 data. Using en-face images obtained by OCT, the central 10-degree visual field was estimated, and a high degree of concordance with actual HFA10-2 data was obtained. This method may be useful for detecting functional abnormalities based on structural abnormalities.

Depth‐resolved variations in visibility of retinal nerve fibre bundles across the retina in enface OCT images of healthy eyes

“…While often included as a limitation, most studies using enface OCT imaging in glaucoma have not accounted for differences between retinal regions, nor between different eyes. [10][11][12]14 According to our data, the conventional 50 µm thick slab might only be inclusive of all RNFBs in a limited number of retinal regions. In the nasal macula, the superior/inferior central retina and the ONH region, information regarding the status of bundles might be overlooked by this approach (Figure 3), as shown in one previous report.…”

“…Most previous work has collapsed a volume of a fixed retinal thickness into a single 2D slab image to explore glaucomatous defects. [10][11][12]14 Among the earliest, Hood et al 11 assessed loss of reflectivity of the central retina using fixed thickness slabs obtained by averaging intensity values up to 52 µm below the ILM (roughly corresponding to the average of 14 slabs in our study). Their axial depth was a compromise between a small thicknessable to show local changes of reflectivityand a large enough number of pixels to reduce noise.…”

“…Additionally, our analysis is the first to consider single slabs of the maximum axial resolution as the unit of assessment of RNFB reflectivity. Most previous work has collapsed a volume of a fixed retinal thickness into a single 2D slab image to explore glaucomatous defects 10–12,14 . Among the earliest, Hood et al 11 assessed loss of reflectivity of the central retina using fixed thickness slabs obtained by averaging intensity values up to 52 µm below the ILM (roughly corresponding to the average of 14 slabs in our study).…”

“…Defects are likely to be induced from a combination of primary loss of reflectivity, caused by cytoskeletal disruption, 9,13 and thinning of the RNFL which leads to the inclusion of deeper hypo‐reflective retinal layers in the slab 11 . Previous studies of enface OCT imaging in glaucoma showed a strong correlation with circumpapillary RNFL defects, 10 and excellent agreement between residual RNFBs and preserved visual function in the macula 12,14 …”

“…11 Previous studies of enface OCT imaging in glaucoma showed a strong correlation with circumpapillary RNFL defects, 10 and excellent agreement between residual RNFBs and preserved visual function in the macula. 12,14 It has been suggested that disruption of RNFB reflectivity may be measurable earlier than RNFL thinning in glaucoma, 8,15 making this approach encouraging for earlier identification of defects. Further, clinicians are encouraged to confirm a suspected structural lesion in the functional domain to increase confidence in diagnosis.…”