Optical imaging of the chorioretinal vasculature in the living human eye

Abstract

Detailed visualization of microvascular changes in the human retina is clinically limited by the capabilities of angiography imaging, a 2D fundus photograph that requires an intravenous injection of fluorescent dye. Whereas current angiography methods enable visualization of some retinal capillary detail, they do not adequately reveal the choriocapillaris or other microvascular features beneath the retina. We have developed a noninvasive microvascular imaging technique called phase-variance optical coherence tomography (pvOCT), which identifies vasculature three dimensionally through analysis of data acquired with OCT systems. The pvOCT imaging method is not only capable of generating capillary perfusion maps for the retina, but it can also use the 3D capabilities to segment the data in depth to isolate vasculature in different layers of the retina and choroid. This paper demonstrates some of the capabilities of pvOCT imaging of the anterior layers of choroidal vasculature of a healthy normal eye as well as of eyes with geographic atrophy (GA) secondary to age-related macular degeneration. The pvOCT data presented permit digital segmentation to produce 2D depth-resolved images of the retinal vasculature, the choriocapillaris, and the vessels in Sattler's and Haller's layers. Comparisons are presented between en face projections of pvOCT data within the superficial choroid and clinical angiography images for regions of GA. Abnormalities and vascular dropout observed within the choriocapillaris for pvOCT are compared with regional GA progression. The capability of pvOCT imaging of the microvasculature of the choriocapillaris and the anterior choroidal vasculature has the potential to become a unique tool to evaluate therapies and understand the underlying mechanisms of age-related macular degeneration progression.

Keywords: Fourier-domain optical coherence tomography; ocular circulation; ocular vasculature; ophthalmic imaging; optical angiography.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Fig. 1.

Comparison of fundus photographs and pvOCT en face projection images for a scanning area of 1.5 × 3 mm2. Selected regions of FA (A) and ICGA (B) of a healthy volunteer from large field photographs (Fig. S2). The cross-sectional composite depth image (C) of OCT intensity (gray) and pvOCT (red) is presented for the retinal location identified by the yellow dashed line in A. En face maximum projection of pvOCT data segmented from retinal layers and at a depth of 5.2 µm underneath Bruch’s membrane, to present the features of the retinal vessels (D) and the choriocapillaris (CC) (E), respectively. En face projection of pvOCT for retinal vasculature (D) demonstrates comparable microvascular visualization to FA (A). Inverting intensity of minimum projection (F) of the pvOCT data segmented a depth slice of 35 µm thickness, located ∼50 µm below the RPE. This inverse image visualizes large feeder vessels in the choroid, where there is usually non–phase-variance signal inside of these vessels shown by green arrowheads in C. The color-coded images (G and H) are combined from retinal vessels (D, red), CC (E, green), and deep choroidal vessels (F, blue). (Scale bars, 200 µm.)

Fig. 2.

Imaging of choriocapillaris of the peripheral retina for a scanning area of 3 × 1.5 mm2 with additional identification of arterioles and venules in the choroid. Color fundus photograph (A) zooming in the region of the yellow dotted rectangle in the full-size image (Fig. S5A). Maximum projection of pvOCT data from retinal layers (B) and CC (C) and Sattler’s layer (D). The yellow dashed line in D is shown in cross-sectional views in Fig. 5 A–D. Segmented depths for these images are positioned 9 µm underneath Bruch’s membrane for the CC and 16 µm for Sattler’s layer. The color-coded image (E) is combined from retinal capillaries (B, red) and CC (C, green) as well as the image (F) consisting of CC (C, green) and Sattler’s layer (D, magenta). (Scale bars, 200 µm.)

Fig. 3.

Small field of view images (1.5 × 1.5 mm2) at the atrophic region of the patient with GA. Selected regions of color fundus (A), red-free (B), AF (C), and the early venous stage of FA (D) images from the location identified in the green square (Fig. S5). The yellow dashed line in D is shown in cross-sectional views in Fig. 6 E–H. En face projection images for retinal vasculature (E), vessels in the inner choroidal layer (F), and large choroidal vessels (G) from pvOCT data (Movie S3). A composite image (H) of three depth layers (retina, red; inner choroid, green; deep choroid, blue). (Scale bars, 200 µm.)

Fig. 4.

Identification of abnormality in CC at edges of GA for a scanning area of 3 × 1.5 mm2. Selected regions of color fundus photograph (A), AF (B), and an early phase of FA (C) from the location identified by the purple dotted rectangle (Fig. S5). An en face image (D) from pvOCT data shows fine normal retinal capillaries (yellow arrows) at the edge of GA, which are not as well visualized with FA (C). A CC perfusion map (E) from pvOCT illustrates small focal areas of CC dropout near the inferior edge of GA, which correlates with the areas of irregular choroidal filling on FA (C). A color composite image (F) of retinal circulation (D, red) and CC (E, green) from pvOCT. (Scale bars, 200 µm.)

Fig. 5.

Depth-resolved image of capillaries in the retina and the choroid. A standard OCT cross-sectional scan (A) averaging three B-scans within a BM-scan (multiple B-scans at the same position) and a composite image (B) of the OCT B-scan (gray) and pvOCT image (red). The retinal position of Fig. 5 A–D is identified on the vertical yellow dashed line in Fig. 2D. Magnified version of OCT intensity (C) and composite imaging (D) cropped from the location indicated at the yellow dashed rectangles in A and B. An OCT B-scan (E) and a pvOCT red-coded image (F) of the patient with GA, with a retinal scan location as identified on the horizontal yellow dashed line Fig. 3D. Cropped images (G and H) from the location of the yellow dashed rectangles in E and F. As indicated by the blue arrow (Fig. 5H), the phase-variance signals by microcirculation disappeared underneath Bruch’s membrane due to CC atrophy, whereas motion contrast signals by blood perfusion (remaining CC) exist in the region pointed by the yellow arrowhead. Volumetric imaging of color-coded B-scans from above two datasets is included (Movies S5 and S6). Horizontal image size for A, B, E, and F: 1.5 mm. (Scale bars, 200 µm.)

Fig. 6.

Schematic of the OCT system. A fiber-based Michelson interferometer was used with an infrared light source having 855-nm central wavelength. The power of the collimated light at the pupil of the eye is 700 µW, below the safety limit of the American National Standards Institute. The light is focused onto the posterior eye, and the back-reflected light from the sample and the reference mirror are combined in the spectrometer portion of the interferometer, which uses a line-scan camera to acquire spectral fringes of the combined light. The Fourier transform of a single spectral fringe pattern generates depth information of a single point, both an intensity profile (axial scan or A-scan) and its phase profile. The X–Y scanner permits acquisition of 3D datasets over defined patches of retina.