Figure 1.
The adaptive optics scanning laser ophthalmoscopy (AO-SLO) system comprises 4 primary optical subsystems: the AO subsystem including the wavefront sensor, the high-resolution confocal SLO imaging subsystem, the wide-field imaging subsystem, and the pupil observation subsystem for initial alignment of the subject's pupil with the optical axis of the AO-SLO system by adjusting the chin rest. The AO subsystem incorporates a liquid-crystal spatial light modulator (LC-SLM) based on liquid crystal-on-silicon (LCOS) technology. The LC-SLM and wavefront sensor are controlled using a custom software to reduce the wavefront errors. The results of high-resolution imaging are linked to the results of the wide-field imaging subsystem, which are obtained by the line-scan SLO system.
Figure 2.
Images of the normal retinal nerve fiber layer (RNFL) using AO-SLO compared with red-free fundus photography and red-free SLO.
A, Wide-field montage of high-resolution AO-SLO images (1.5°×1.5°) within a 30° arc from the foveal center. AO-SLO images show many hyperreflective bundles in the RNFL. The hyperreflective bundles above and below the fovea appear in an arched shape from the temporal periphery on either side of a horizontal dividing line to the optic disc. The hyperreflective bundles on the nasal side of the fovea arch from the fovea to the optic disc. Small white boxes (a–e) indicate the area of high-magnification AO-SLO images in Fig. 5. B, Magnified view of the area outlined in large white box in A showing individual hyperreflective bundles. C, Red-free SLO image of the same eye obtained with an F-10 (NIDEK, Gamagori, Japan). The hyperreflective bundles on AO-SLO (A, B) correspond with the direction of the striations on the SLO red-free image. D, Magnified blue-channel fundus photography image of the area inside the box in C. E, Magnified red-free SLO image (HRA2; Heidelberg Engineering, Heidelberg, Germany) of the area inside the box in C. F, Magnified red-free SLO image (F-10) of the area inside the box in C. G, Magnified AO-SLO image of the area inside the box in C. The resolution and contrast of the bundles are much higher in the AO-SLO images (G) than in red-free fundus photography (D) or the red-free SLO images (E, F).
Figure 3.
Measurement of the width of hyper-reflective bundles.
To measure the width of individual hyper-reflective bundles, 3–8 bundles were randomly chosen from 1 AO-SLO image (1.5°×1.5°). The digital caliper tool was used to measure the width at a minimum of 3 points in 1 bundle by 2 independent experienced graders (i.e., the bundle width of this area was defined as the mean width of 18 points).
Figure 4.
Comparison of the number of peaks of plot profile between AO-on images and AO-off images.
A, B. AO-on image focused on the RNFL. Mean gray value of images was plotted (190 pixel width) along a line that vertically crosses the hyper-reflective bundles. C. Plot profile of gray value along the yellow line of B. The number of peaks, which was defined as more than 20 gray values compared to the neighboring baseline, was 8. D, E. AO-off image of same area. F. Plot profile of gray value along the yellow line of E. The number of peaks was zero.
Figure 5.
High-magnification RNFL images using AO-SLO of the area indicated by white boxes in Fig. 2
(a–e in Fig. 2 corresponds to A–E in Fig. 5). A, Image near the optic disc indicated by a in Fig. 2. B, Image 2 mm above the fovea. C, D Images at the temporal raphe. Bridges can be seen among the hyperreflective bundles (arrows). E, Image of the area around the papillomacular bundle. The dark lines among the hyperreflective bundles are narrower around the optic disc (A) than at the temporal raphe (C, D). The hyperreflective bundles on the nasal side of the fovea (E) are thinner than those above (B) or below the fovea. Scale bar = 100 µm (A–E).
Figure 6.
Mean widths of the hyperreflective bundles at various distances from the optic disc.
(Top,Middle) Mean widths of the hyperreflective bundles in normal eyes at 1∼6 mm from the optic disc. Measurement points are 1 mm, 2 mm, 3 mm, 4 mm, 5 mm, and 6 mm from the optic disc. There are no significant differences among the bundle widths at these distances. (Bottom) RNFL thickness was measured by SD-OCT in normal eyes at the corresponding area on AO-SLO measurement points. Note that the RNFL thickness on SD-OCT deceases in proportion as distances from the optic disc increase, whereas the hyperrefrective bundle width remains constant. The error bars are SDs of 20 normal eyes.
Figure 7.
Mean widths of the hyperreflective bundles at the same distance around the optic disc.
(Top, Middle) Mean widths of the hyperreflective bundles in normal eyes at a diameter of 3.4 mm centered on the optic disc. The hyperreflective bundles on the temporal and nasal sides of the optic disc are narrower than those above and below the optic disc; thus, the bundle widths around the optic disc have a double-humped shape. (Bottom) RNFL thickness measured by SD-OCT in normal eyes at a diameter of 3.4 mm centered on the optic disc. Note that the RNFL thickness measured by SD-OCT around the optic disc also has a double-humped shape. The error bars are SDs of 20 normal eyes.