Abstract
Purpose: :
To better understand normal variation of inner retinal layers measured with frequency domain optical coherence tomography (fdOCT), the relationship between several anatomical factors and fdOCT layer thicknesses were examined.
Methods: :
One eye of 128 controls (data supplied by Topcon, Inc.) was tested with fdOCT (3D-OCT 2000, Topcon, Inc). To be included, subjects were required to have BCVA of 20/40 or better, a correction between +3.0 D and -6.0 D, IOP ≤ 21 mmHg, axial length between 22-26 mm, normal clinical exam, and normal SAP (SITA Standard, Zeiss, Inc.; PSD p > 0.05, normal GHT, reliability indices all ≤ 33%). Exclusion criteria included history of ocular disease or a family history of glaucoma. Two 10° "cube" fdOCT scans of the macula and the optic disc (128 B-scans x 512 A-scans) were acquired. A previously validated computer algorithm [1] segmented the images by marking the following borders: vitreous/inner limiting membrane (ILM), retinal nerve fiber layer (RNFL)/retinal ganglion cell (RGC), inner plexiform layer (IPL)/inner nuclear layer (INL), INL/outer plexiform layer (OPL), and Bruch’s membrane (BM)/choroid. The layers measured were: RNFL (ILM to RNFL/RGC), RGC+IPL (RNFL/RGC to RGC/IPL), and the total receptor (TR; the INL/OPL to BM/choroid). Scans with acquisition or algorithm artifacts were rejected. The centers of the macula and disc scans were marked based on both C-face and B-scan images and these cube scans were then coregistered. Pearson coefficients (R) were calculated.
Results: :
The fovea-to-disc angle did not correlate well with the locations of the circumpapillary (cp) RNFL peaks (RSUP=0.09, RINF=0.12) or blood vessels (BVs; RSUP=0.13, RINF=0.14). As suggested,[2] BVs moderately correlated with cpRNFL peak locations (RSUP=0.51, RINF=0.27). The disc radius or the fovea-to-disc distance did not correlate well with the cpRNFL thickness (R=0.02, R=0.18) or the eccentricity of the peak of the macular RGC+IPL thickness (R=0.13, R=0.12). The macular RGC+IPL peak height moderately correlated with the mean cpRNFL thickness (R=0.37), while the mean macular TR thickness did not correlate well with the mean cpRNFL thickness (R=0.20).
Conclusions: :
Blood vessel locations may have potential in reducing variability. The potential of other anatomical factors in this study seems limited. 1. Yang Q, et al., Opt. Exp. 2010; 2. Hood DC, et al. J. Glaucoma 2010.
Keywords: imaging/image analysis: clinical • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • retina: proximal (bipolar, amacrine, and ganglion cells)