Role of OCT Angiography in Neovascular AMD

Benefits in diagnosis and clinic workflow.


The high-resolution imaging offered by OCT has revolutionized the management of AMD. Recently, OCT angiography (OCTA) was introduced to the field of retina and many retina specialists are beginning to acquire the technology. In this article, we review some of the benefits and limitations of using OCTA in AMD and its associated variants.


OCTA utilizes advances in image acquisition speed and resolution to generate detailed vascular maps in 3 dimensions. By obtaining images in rapid succession and evaluating them for differences, software within OCT devices is able to identify blood flow in vasculature.

These images are depth encoded, resulting in a 3-dimensional volume. This image can be scanned through and examined for vascular abnormalities at all retinal levels.1


Previously, the gold standard for imaging choroidal neovascular membranes (CNVM) was fluorescein angiography (FA). OCT has allowed us to indirectly evaluate for CNVMs through the presence of fluid within or underneath the retina. The introduction of OCTA technology has allowed retina specialists to image CNVMs without the invasive and time-consuming process of FA.1 By scrolling through the OCTA volume, CNVMs can be localized in the macula both in distance from the fovea and in retinal depth.(Figure 1; Video).

Video. OCTA demonstrating a retinal angiomatous proliferation (RAP) lesion. Bottom right-hand image shows an OCT B-scan demonstrating flow (yellow) signal traversing the depth of the entire retina next to the blue line. As the segmentation selection lines (red dashed lines) go from the inner retinal layers through the outer retinal layer, the en face angiography image in the upper-right hand image shows the structure of the RAP lesion traversing the entire retina in the crosshairs of the blue and green lines. The blue line corresponds to the blue localizer line on the OCT B-scan and the green line is the location of the OCT-B-scan in the macular cube. Two other RAP lesions are noted as well in the en face angiographic images superior and inferiorly as well but outside the selected B-scan.

OCTA Showing RAP Lesion from Pentavision on Vimeo.

Figure 1. OCTA demonstrating a type 2 CNVM above the RPE. The OCT B scan in the lower right-hand corner shows proper segmentation of the CNVM between the dotted red lines. The yellow signal within the CNVM identifies flow, which is then processed into the image in the upper right-hand corner showing the neovascularization within the membrane.

We find OCTA to be incredibly useful in evaluating dry AMD patients suspected of converting to wet AMD. These patients can present with new symptoms and subtle outer retinal changes on OCT that are questionable. In these patients, we would traditionally have ordered an FA to clarify if there truly was neovascularization present. Now our photographers are instructed to automatically acquire an OCTA along with the OCT on all patients suspected of conversion. This has greatly improved our clinic flow time, as patients aren’t going back and forth to photography for additional testing.


The 3-dimensional aspect of OCTA images helps to distinguish type 1 and type 2 CNVMs based on their presence below or above the RPE, respectively.1 This depth information can also be useful for identifying type 3 CNVMs, also known as retinal angiomatous proliferation (RAP) lesions, by scrolling through the volume cube and following the RAP lesion as it dives deeper into the outer retina and choriocapillaris (Figure 2).2

Figure 2. OCTA showing a retinal angiomatous proliferation (RAP) lesion. OCT B scan demonstrating flow (yellow) signal traversing the depth of the entire retina within the orange rectangle (A); OCTA en face images show the RAP lesion within the orange circle in the superficial, deep, and avascular segments of the retina, respectively (B-D).

We have attempted to use OCTA in the diagnosis of polypoidal choroidal vasculopathy (PCV). In our experience, OCTA often misses the polyps within the choroidal space in these patients. We still often have to rely on traditional angiography (FA and ICG) to diagnosis patients with PCV. One theory for the lack of effective visualization is the blood flow within polyps may be lower than the surrounding vasculature. This low flow is likely below current OCTA ability for detection.3

Central serous retinopathy is another diagnosis that can often mimic exudative AMD. The diagnosis of secondary neovascularization can be especially challenging because fluorescein leakage occurs regardless of whether neovascularization is present. OCTA is not affected by the underlying leakage, and thus can directly image neovascularization.4 The ability to image deep vasculature in eyes with leakage is a key benefit.


OCTA has made significant inroads in our management of AMD patients, but there are still significant limitations that we encounter. Clinically, there are 4 in particular that we encounter the most: ocular surface irregularities, motion artifact, segmentation errors, and projection artifact.

Ocular surface irregularities: Irregularities of the ocular surface, such as dry eye or corneal irregularities, can compromise scan quality and introduce noise into the final image.

Motion artifact: OCTA requires 2 scans, acquired in rapid succession, for comparison. If there is significant eye movement between the 2 scans, noise can be introduced into the final OCTA image. AMD patients with low vision have difficulty with fixation and their subtle saccadic eye movements impair the ability to get quality scans. As OCTA vendors improve their eye tracking components, we anticipate this to be less of an issue.

Segmentation errors: All OCTA devices divide retinal layers into inner, outer, and avascular segments. There are often other segmentations, such as vitreoretinal interface and choriocapillaris. In pathologic eyes, the devices can make mistakes in tracing the retinal layers, resulting in the omission of important details when viewing the OCTA report. We often manually scroll through OCTA volumes to make sure nothing was missed.

Projection artifact: OCTA is susceptible to an artifact where superficial retinal vessels are seen in deeper retinal structures. With our fellows, this often causes confusion and misdiagnosis of CNVMs that are not actually present. Currently, we examine the superficial layers and compare them to deeper layers to see if there are similar patterns. If so, we ignore those vessels. We also use the OCT B scan with flow overlay to see if superficial vessels are projecting deeper in the retina. The presence of this artifact is challenging, and makes identification of small CNVMs difficult (Figure 3).

Figure 3. OCTA demonstrating projection artifact. In image B, the OCTA en face image is suggestive of a large type 2 CNVM. However, the OCTA B scan in image C shows lack of yellow signal, which is indicative the absence of flow within the concerning membrane that is being segmented between the red dotted lines. Furthermore, comparison of the image A to image B in the upper right shows similarities with the superficial retinal vasculature, thus confirming projection artifact.

Awareness of these limitations has greatly improved our ability to read OCTA images.


OCTA is an exciting development for the management of AMD patients. The technology is still in its infancy, but we have found benefit in our clinical setting both from a diagnostic and workflow stand point. RP


  1. Spaide RF, Klancnik JM, Cooney MJ. Retinal vascular layers imaged by fluorescein angiography and optical coherence tomography angiography. JAMA Ophthalmol. 2015;133 (1),45-50.
  2. Tsai AS, Cheung N, Gan AT, et al. Retinal angiomatous proliferation. Surv Ophthalmol. 2017;62(4):462-492.
  3. Wang M, Zhou Y, Gao SS, et al. Evaluating polypoidal choroidal vasculopathy with optical coherence tomography angiography. Invest Ophthalmol Vis Sci. 2016;57(9):526-532.
  4. Bonini Filho MA, Talisa E, Ferrara D, et al. Association of choroidal neovascularization and central serous chorioretinopathy with optical coherence tomography angiography. JAMA Ophthalmol. 2015;133(8):899-906.