Peripapillary nerve fiber layer thickness profile determined with high speed, ultrahigh resolution optical coherence tomography high-density scanning

Abstract

Purpose: To determine the retinal nerve fiber layer (RNFL) thickness profile in the peripapillary region of healthy eyes.

Methods: Three-dimensional, Fourier/spectral domain optical coherence tomography (OCT) data were obtained as raster scan data (512 x 180 axial scans in a 6 x 6-mm region centered on the optic nerve head [ONH]) with high-speed, ultrahigh-resolution OCT (hsUHR-OCT) from 12 healthy subjects. RNFL thickness was measured on this three-dimensional data set with an in-house software program. The disc margin was defined subjectively in each image and RNFL thickness profiles relative to distance from the disc center were computed for quadrants and clock hours. A mixed-effects model was used to characterize the slope of the profiles.

Results: Thickness profiles in the superior, inferior, and temporal quadrants showed an initial increase in RNFL thickness, an area of peak thickness, and a linear decrease as radial distance from the disc center increased. The nasal quadrant showed a constant linear decay without the initial RNFL thickening. A mixed-effects model showed that the slopes of the inferior, superior, and nasal quadrants differed significantly from the temporal slope (P = 0.0012, P = 0.0003, and P = 0.0004, respectively).

Conclusions: RNFL thickness is generally inversely related to the distance from the ONH center in the peripapillary region of healthy subjects, as determined by hsUHR-OCT. However, several areas showed an initial increase in RNFL, followed by a peak and a gradual decrease.

Figure 1

(A) hsUHR-OCT en face fundus image. (B) RNFL thickness map. Red: increased RNFL thickness; blue: decreased thickness. The ONH region (circle) was excluded from RNFL analysis.

Figure 2

Quadrant (left) and clock hour (right) mean (solid line) and 95% confidence intervals (dotted lines) of nerve fiber layer thickness profiles in all subjects. Vertical dashed line: the location of the conventional OCT scan circle (1.7 mm from the center of the ONH).

Figure 3

A comparison of RNFL thickness profiles as measured from the center of the ONH (left) and from the disc margin (right).

Figure 4

Average nerve fiber layer quadrant thickness for all subject quadrants, after mean filtering.

Figure 5

Average variance in nerve fiber layer thickness for all subjects after mean filtering. Superior and inferior quadrants are regions with more blood vessels, and a higher variability can be seen in these quadrants than in nasal and temporal quadrants. Vertical dashed line: location of the conventional OCT scan. The minimum total variance was observed at 1.62 mm from the disc center, which was close to the standard RNFL sampling location (1.7 mm; vertical dotted line) of the commercial OCT unit.

Figure 6

Algorithm failure. RNFL thickness map, where red marks the areas of the thicker RNFL and blue the thinner areas (left). The circle covers the ONH. A cross-sectional scan (right) was taken along the red line just under the disc margin. Red arrows: RNFL border detection algorithm failure toward the edge of the OCT image, where the signal was weaker. Blue arrows: an area just outside the ONH border where several blood vessels were located and the RNFL was thinner.