OCT Angiography

January 16, 2016
OCT Angiography

In my 27 years in ophthalmology,  there have been a few watershed moments that stand out. In the early nineties, it was the transition to digital imaging, followed quickly by ICG angiography. In the late nineties, I was fortunate to witness the early advent of OCT, and was in the room when the results of the Marina and Anchor studies, which brought anti-VEGF therapy into the office, was announced.

At the American Academy of Ophthalmology Annual Meeting in Chicago in 2014, you would be hard pressed to find much information on OCT angiography (OCTA). At the meeting in Las Vegas this past Fall, OCTA was the topic of most conversations. At the time of this writing, the only FDA approved OCTA is the Zeiss Angioplex. I have used the Zeiss Angioplex system in clinic since the beginning of this year, and wanted to share some of my preliminary thoughts.

First, I have to keep reminding myself that we are in the pioneering stage of OCTA, and therefore rules have yet to be established. There are no set protocols for scanning; it will take some time to create those, and will vary by practice. Second, because we are in the pioneering stage, we don’t really know what we are seeing. As was the case in the early days of OCT, pathology presents itself differently than what we’ve become used to. We have to not only identify what is normal, but also come up with accepted nomenclature for everything we see on the images.

Using en-face visualization of the layers in the retina, the OCTA detects flow in the retinal vessels, to the capillary level, by identifying movement in the posterior pole. The algorithm assigns white color to movement and black to non-movement, which gives the image the appearance of a fluorescein angiogram. The Angioplex system allows for the option of color depth encoding, which assigns color to the vessels according to their depth in the retinal or sub retinal layers. The easiest way to describe OCTA is this:
Standard OCT displays anatomy, OCTA displays function. 

color superficial
Normal retina. The color depth encoding image assigns color according to the depth of the vessels. Using en-face imaging, you can visualize vessels at the superficial, deep, avascular, choriocapillaris and choroidal level

Angioplex is built on the foundation of the Zeiss Cirrus OCT, which also allows for visualization of the en-face images of the retina. Coupled with their spectral domain OCT, I am able to produce line scans, raster scans, volumetric OCT, etc, along with the accompanying OCTA images. The movie mode displays each layer of the en-face scan IN OCTA mode, so I can find subtle changes in the angiogram and correlate it with the OCT scan.

There are some restrictions to the technology at this point. OCTA can not detect staining or pooling of fluid. For instance, if a patient presents with sub retinal fluid secondary to AMD or CSCR, but there is no active leakage, the OCTA may present as normal. The other major restriction is the lack of wide angle imaging. Presently, the Angioplex system allows for 3mmx3mm or 6mmx6mm, although future applications may include 8×8 and 12×12.
Macleod_Billy__187013_19870603_Male_Angiography 6x6 mm_20160107081517_OD_20160107132429 - Retina.AngiographyEnface  Macleod_Billy__187013_19870603_Male_Angiography 6x6 mm_20160107081551_OD_20160107132414 - Superficial.AngiographyEnface
      6 x 6 scan of non perfusion                 same patient, scanned with 3 x 3

The downside to being “ahead of the curve” is that the network of users is limited. The advantage is that you can be a part of establishing protocol for new technology; protocols that become standard across the country.
As I become more familiar with the technology, I will continue to post blogs updating my progress. Here are some images from the past week:

Kelly_Richard__brvo- Deep.AngiographyEnface   Macleod_Billy__187013_19870603_Male_Angiography 6x6 mm_20160107081419_OD_20160107132327 - Retina Depth Encoded.AngiographyEnface
          BRVO at the deep retina level            NVD on color depth encoding

Weber_William__rap - Retina Depth Encoded.AngiographyEnface   Burns_Chad__187610_19690410_Male_Angiography 3x3 mm_20160113094325_OD_20160114154915 - Retina Depth Encoded.AngiographyEnface
                                              CNV on color depth encoding

Webster_Daniel__187587_19490309_Male_Angiography 3x3 mm_20160111092102_OS_20160111164528 - Superficial.AngiographyEnface  Webster_Daniel__187587_19490309_Male_Angiography 3x3 mm_20160111092102_OS_20160111164534 - Deep.AngiographyEnface   Webster_Daniel__187587_19490309_Male_Angiography 3x3 mm_20160111092102_OS_20160111131548 - Retina Depth Encoded.AngiographyEnface
CNV on superficial retina, choriocapillaris and color depth encoding

ICG angiography

ICG (Indocyanine Green) angiography is a tool used by retinal specialists to primarily visualize choroidal blood vessels. The dye is used either stand alone or as an adjunct to fluorescein angiography (FA). It is important to understand the difference between ICG and fluorescein.
Unlike fluorescein, ICG is a large molecule dye and has a 98% binding to serum, which prevents it from leaking from normally porous choroidal vessels. ICG molecules are excited by a higher wavelength of light than fluorescein. The emission of excited light is at 835 nm, which essentially eliminates the absorption of light in the retinal pigment epithelium, allowing, in turn, the visualization of choroidal vessels.
ICG comes in powder form, and requires reconstitution with a dilutant; typically an aqueous solvent. The dosage is dependent on the amount of dilutant used to reconstitute, and tends to vary from practice to practice. Once reconstituted, the dye is only stable for 10 hours.
Choroidal blood vessels are one of the fastest vascular flow complexes in the body. Imaging dye as it fills these vessels can be a challenge. Not only is it fast, but there is no “normative” vessel pattern, as there is with retinal vessels; so we typically can’t image an arterial, venous or transit phase.
We can’t think of ICG as we do with fluorescein in regards to sub retinal neovascularization. In FA, we look for leakage patterns to determine the type or extent of a neovascular lesion. ICG allows us to image the SOURCE of the lesion. With ICG, one can differentiate between a choroidal vessel source (CNV) or a retinal vessel source (RAP lesion). With the elimination of laser treatment for wet AMD, the days of looking for a late “hotspot” on ICG are over. Feeder vessels are also identifiable on ICG angiography, and diagnosing the source of the lesion allows the physician a better prognosis on extent or length of treatment.
ICG and FA can be performed simultaneously or separately. Mixing a usual dosage of fluorescein with a usual dosage of ICG allows for one injection and  the imager to switch back and forth between FA and ICG filters to image both retinal and choroidal vascular layers. Separating the dyes by a stopcock or valve is another option. After the initial early phase of both angiograms, the process is the same. Here is the protocols I’ve used:

Mixing dyes:

Draw up the usual dosage of fluorescein in one syringe, and the usual dosage of ICG in another, larger syringe. Inject the fluorescein into the ICG syringe.
Connect the syringe to the IV tubing and inject as usual. At this point the imager has only one choice for early images: ICG or FA, as both dyes reach the eye at the same time. Once early images are captured, switch filters and RE FOCUS on the proper layer. Remember that choroidal and retinal vessels are at different depths from each other.
Continue the angiograms, switching filters, flash and software screens.

Separate dyes:

Draw up the usual dosage of fluorescein in one syringe, and the usual dosage of ICG in another syringe. Attach the syringes to a 3-way stopcock and attach stopcock to the IV tubing. Inject one of the dyes (It doesn’t matter which one you inject first), and capture the early images. Switch filters, flash and software, and inject the other syringe.
Continue going back and forth per your practice protocol.

My 2 cents on timing:
As we are looking for the SOURCE of sub retinal lesions, ICG is most useful in the early phase of the angiogram. Fluorescein is also helpful in the early phase, prior to the start of the leakage of dye. Mid to late phase of the fluorescein will reveal leakage. Therefore, both ICG and fluorescein late images are taken at 5 minutes.


Foveal scanning in epiretinal membrane

Finding a foveal depression on an OCT can be challenging at times. In the presence of an epiretinal membrane (ERM), it can be like trying to find Waldo.



Although the pathogenesis of primary, idiopathic ERM is relatively unknown, the effect of ERM is certainly known. In advanced stages of ERM, there can be significant traction on the retinal layers, and result in edema. Patients complain of decrease vision and metamorphopsia, and progression of ERM may result in a macular hole.

Because of the tension that ERM inflicts on retinal tissue, the normal foveal depression becomes vaulted and flush with the level of the internal limiting membrane (ILM). This makes it difficult to identify the fovea and provide a proper scan of the retina.

The vaulted fovea will result in a peaked appearance on OCT and the apex of this peak is the center of the fovea. By moving the scan line, the peak will taper and fall as you move through the fovea. It’s important to move the scan until the image shows the tapered end of the fovea where it meets the top (ILM) of the retinal tissue.

erm GreenleafJ erm 1 PerryEThe two scans above are typical foveal peaking under an ERM

erm with rounded fovea
The scan above shows a rounded peak, requiring the imager to move the scan to find a more tapered peak

In the scans below, a simple horizontal movement of the scan reveals a more peaked appearance to the fovea:erm fovea peak round
Note the rounded appearance to the peak of the fovea, as well as the top of the peak not reaching the top of the retinal tissue

erm fovea peak sharp
By moving the scan, the foveal peak appears more tapered and reaches the top of the retinal tissue


Optimizing your OCT scans

Optical Coherence Tomography utilizes infrared light that renders an in vivo cross sectional view of the retina. Therefore, anything that affects the light pathway WILL affect your image. Common artifact issues can be attributed to cataracts, vitreous opacities, and corneal media interference.
The act of performing application tonometry and the drying effect of some eyedrops, such as anesthetic and dilating drops, may disrupt the tear film and epithelium. This will scatter light entering the eye, and more importantly, reduce the focused light coming out of the eye and back to the sensor on the OCT. The resulting image will have a “snowy” or “static” appearance throughout the vitreous, and the layers of the retina will not be as evident.
One way to reduce this problem is to scan the patient prior to the technical workup. This will allow the imager to scan through a cornea that has not been subjected to drying effects.
Here is an example of a patient that was scanned prior to begin worked up and after being worked up.


            Before work up


              After work up    



Full Depth Imaging on OCT

OCT tip of the month- Full Depth Imaging(FDI), or Combined Depth Imaging(CDI).

Posterior OCT has a small window of “focus”- in other words, the depth of focus is usually from the pre retinal vitreous face to the RPE.

When we use the Enhance Depth Imaging (EDI) utility on the OCT, it moves the window of focus further posteriorly- from approximately the inner plexiform layer (IPL) to well into the choroid, and even to the sclera. This allows for much better visualization of the choriocapillaris and choroid, and allows for choroidal thickness measurement. The downside is that we lose detail of the neurosensory retinal layers- the nerve fiber layer (NFL) to the outer plexiform layer (OPL). FDI, or CDI, allows us to combine both modes of OCT to produce an image that is detailed from NFL to sclera.

Using the Heidelberg Spectralis, it’s easy to combine both techniques to get Full Depth Imaging.

  1. Line up a scan as usual.
  2. Set the ART on the screen to 100 frames

IMG_3237arrow copy

  1. Press the ART/Sensitivity knob on the touch screen


  1. Once the ART images have reached the halfway point, click on EDI


  1. Allow the ART frames to reach 100 in the EDI mode


The resulting image will be a combination of standard OCT and EDI OCT

oct regular scan LEBLANCS

Standard OCT




Combined, or Full Depth Imaging OCT