Purpose:
The reflectivity of the retinal nerve fiber layer (RNFL) has been shown to decrease in glaucoma (Van der Schoot J, et al. IOVS 2010;51:ARVO E-Abstract 212). Our aim was to assess local RNFL reflectivity in glaucoma as determined by selective integration of OCT images.
Methods:
10 normal eyes and 8 glaucomatous eyes (1 eye per subject) were scanned with a Spectralis OCT (Heidelberg Egineering, Germany; 20° FOV centered on papilla, 512 A-lines per B-scan, 193 B-scan per volume, 5 times averaging). They were also imaged by a GDx (Carl Zeiss Meditec, Dublin, CA; ECC mode was used). An automatic segmentation method was trained on two manually segmented B-scans of each normal eye (Vermeer KA, et al. IOVS 2010;51:ARVO E-Abstract 219). The algorithm then automatically segmented the RNFL of the glaucomatous eyes. Finally, RNFL reflectivity maps were calculated by averaging the OCT signals over the segmented RNFL per A-line. For evaluation, the GDx images were first graded on local and diffuse loss. Then, the reflectivity maps were assessed for irregular appearance. Both scores were compared.
Results:
Examples of RNFL reflectivity maps are shown in Figure 1.
In normal eyes, the reflectivity maps are rather uniform, in contrast with thickness images that show distinct arcuate RNFL bundles. Almost all defects that were found in the GDx images were also located in the reflectivity maps and often easier to assess. In some cases, the reflectivity maps were more detailed, showing for example multiple local defects instead of single ones, a large wedge instead of diffuse loss or local defects in addition to diffuse loss. Two local defects were not readily visible in the reflectivity map.
Conclusions:
Reflectivity maps provide additional data of the RNFL morphology. These maps are mostly consistent with GDx data but seem to provide more details. They are also far easier to read due to their uniform appearance in normal eyes. Reflectivity maps may thus provide a novel diagnostic tool for glaucoma assessment.
Keywords: image processing • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • optical properties