Color Editing and 3D Volume Rendering of OCT-angiography Images

Abstract

Purpose

To describe an innovative imaging technique of color editing and 3D volume rendering of optical coherence tomography-angiography (OCT-A) images.

Methods

Each vascular layer acquired by OCT-A (AngioVue, Optovue) is colored and edited with the GNU Image Manipulation Program. Then a 3D volume rendering of the assessed retinal area is possible, using Adobe After Effect software (Adobe Systems).

Results

The combined color editing and 3D volume rendering of optical coherence tomography angiography images allow the creation of a short movie displaying the 3D image from chosen angles.

Conclusions

Image editing and 3D volume rendering is a rather simple technique, which, besides its aesthetic prerequisites, also facilitates simultaneous, distinct visualization of the segmented retinal layers.

Introduction

Optical coherence tomography-angiography (OCT-A) is a novel and noninvasive imaging tool that allows the visualization of retinal microvasculature by detecting intravascular blood flow without any dye injection. The split spectrum amplitude decorrelation algorithm was used to distinguish blood flow from static tissue (1).

This new type of imaging is depth-enhanced, allowing a detailed visualization of retinal and choroidal microvascularization at different depth levels. Multiple slabs – generated either manually or automatically, using the software segmentation, or manual segmentation level – are available on all OCT-A machines and prototypes (AngioVue, Optovue; HRA 2, Heidelberg Engineering; DRI OCT Triton, Swept Source OCT, Topcon; OCT Spectral RS-3000 Advance, Nidek SA; Cirrus AngioPlex OCT Angiography, Carl Zeiss Meditec, Inc.).

Method

We have recently devised a technique for color editing and 3D volume rendering of OCT-A images. The images used to exemplify the technique in this article were acquired with AngioVue.

The OCT-A software automatically generates four segmentations: (i) the superficial capillary plexus (SCP); (ii) the deep capillary plexus (DCP): (iii) the outer retina layer (ORL) and (iv) the choriocapillary layer (CL). The SCP is located between the internal limiting membrane and the posterior part of the inner plexiform layer, while the DCP is situated in the inner nuclear layer (2). The ORL, normally avascular, is delimited by the outer boundary of the outer plexiform layer and the retinal pigment epithelium, and the CL is located from the Bruch's membrane to 10-20 μm above.

OCT-A allows the user to save separate images corresponding to each of the four segmentations. Image editing is generated using these separate images, epitomizing the four automatic segmentations. For each image, we have used different layers in the GNU Image Manipulation Program (GIMP). GIMP is a free and open-source raster graphics editor that is used for image editing (http://www.gimp.org, accessed March 5, 2017).

Results

Using the color selection mode we can select and extract various vascular networks at various depths (e.g., SCP, DCP, type 1 or 2 neovascularization) as image layers and then colorize them (Fig. 1). Once the separate images corresponding to all available segmentations have been edited and colorized, a superposition is possible, generating an intuitive image of both the inner retinal microvascularization and the eventual presence of choroidal neovascularization (CNV) (Fig. 2).

Fig. 1

Different slabs of optical coherence tomography-angiography images of a 70-year-old woman with type 1 neovascularization secondary to exudative age-related macular degeneration, showing the superficial plexus (A) the deep capillary plexus (B), and the choroid capillary plexus (C). The colorization of each capillary plexus using GNU Image Manipulation Program software is: in blue, the superficial plexus (A’); in yellow, the deep capillary plexus (B’); and in red, type 1 neovascularization (C’).

Fig. 2

Overlay of the three capillary layers of a 75-year-old woman with type 1 neovascularization secondary to exudative age-related macular degeneration. In blue, the superficial plexus; in yellow, the deep capillary plexus; and in red, type 1 neovascularization.

Furthermore, a 3D volume rendering of the assessed retinal area is possible using Adobe After Effect (AAE) software (Adobe Systems). AAE is an application that generates digital visual effects, motion graphics, and image compositing. The different colored image layers, extracted with GIMP, can be spatially displaced and/or rearranged. Using the “Camera Mode” of the AAE software, we are then able to create a short movie displaying the 3D image from the chosen angles (Fig. 3 and movie in Supplementary Video 1 - Neovascularization. Available online at ophthalmology.pointofcarejournals.com).

Conclusions

Recently, Spaide (3) showed a combined and integrated volume rendering of the retinal vasculature and selected structural abnormalities information derived from OCT.

Fig. 3

Multiples angles of a 3D film made with Adobe After Effect software. In blue, the superficial plexus; in yellow, the deep capillary plexus; and in red, type 1 neovascularization.

Many published studies have highlighted the usefulness of OCT-A imaging in the detection and assessment of macular diseases (4-5-6-7-8). In exudative age-related macular degeneration or retinal vascular diseases, this imaging technique allows both a qualitative and quantitative assessment of capillary modifications, ischemic areas, and neovascularization. Image editing and 3D volume rendering is a rather simple technique, which, besides its aesthetic prerequisites, also facilitates simultaneous, distinct visualization of the segmented retinal layers.