Overview of OCT Angiography (TPA) | California Optometric Association
Quick Links

Latest Updates

Access Member Resources

Children's Vision

Overview of OCT Angiography (TPA)

California Optometry Online Articles Payment Forms Transcripts CE Events

Michael H. Chen ,OD

Dr. Chen graduated from UC Berkeley School of Optometry. He currently resides in the San Francisco Bay Area, and works in a private practice in Santa Clara, CA. He also serves as a clinical research consultant to Carl Zeiss Meditec for advanced product and prototype development, and has worked extensively in the area of OCT angiography. Dr. Chen is a member of the American Optometric Association, the California Optometric Association, and American Academy of Optometry. He is also an active member of his local optometric society. In his free time, he enjoys spending time with friends and family, watching movies, and recording music.

    Patricia J. Sha, OD

            

Patricia Sha, OD earned her Bachelor of Science in Biochemistry and Cell Biology from the University of California, San Diego. Dr. Sha earned her Doctor of Optometry degree from the University of California, Berkeley, School of Optometry, where she was a part of the Beta Sigma Kappa Optometric Honor Society.  After graduation Dr. Sha became an Optometric Consultant at Carl Zeiss Meditec, Inc. As a consultant, Dr. Sha works as a research coordinator on new medical devices relied upon by the optometric community. Dr. Sha also practices private eye care in the Bay Area, CA. Dr. Sha maintains an active role in the optometry community and is a part of the American Academy of Optometry, American Optometric Association, California Optometric Association, and the Santa Clara County Optometric Society, where she acts as the Legislative Events Liaison. Dr. Sha has a continued commitment to furthering the field of optometry in both her research and in private practice.

For a Printable Copy Click HERE

Overview of OCT Angiography (TPA)

INTRODUCTION

Optical coherence tomography (OCT) has revolutionized eye care by providing a fast, noninvasive method to visualize ocular structures in vivo since its development in 1991. It has since become one of the most important ophthalmic diagnostic tools, with an expanding number of applications that aid eye care providers in diagnosing and monitoring eye disease.

Recently, there has been great excitement in the development of a new OCT imaging technique called OCT angiography (OCTA). OCTA isolates microvascular circulation from OCT image data by specialized acquisition and processing techniques. This allows for visualization of blood flow in the retina without the need for intravenous injection of a contrast dye. Until now, clinical visualization of retinal vasculature has been limited to indocyanine green angiography (ICGA) and fluorescein angiography (FA), both of which require contrast dye injections. Despite their great clinical utility, these procedures are not without drawbacks. They are invasive, time consuming, and require a skilled photographer. The injected dye can sometimes cause adverse reactions including nausea, vomiting, and in rare occasions, anaphylaxis possibly leading to death1. Additionally, indocyanine green dye is contraindicated in pregnant patients, or patients with kidney disease2. With the introduction of OCTA, clinicians now have a fast, noninvasive method to visualize retinal microvascular perfusion.

TECHNOLOGY

OCTA works by scanning the same location on the retina multiple times consecutively. As the scanning beams enter the eye, areas with no movement will always reflect light the same way. However, moving structures (such as red blood cells) will reflect light differently between each B-scan. These B-scans are then compared against each other. After eliminating the effects of bulk motion such as head movements or saccades, any residual differences between the scans are thought to be due to blood flow, and hence indicate the location of a perfused blood vessel3. Areas with blood flow appear bright, whereas dark areas indicate no flow (or flow that is too slow to detect). By combining the flow information in all B-scans, an en face image of the microvascular circulation can be generated. These OCTA images can achieve an axial resolution of approximately 5 microns, which gives detailed views of fine capillaries (Figure 1).

Figure 1. OCTA en face image of a healthy eye. Note that the foveal avascular zone can be clearly seen, as can the small retinal capillaries

 

Currently, there are several different OCTA processing methods being used by different companies. These include phase-based techniques (Doppler shift or variance), magnitude-based techniques (speckle variance or decorrelation), or a combination of both4-6. Each method utilizes a different part of the OCT signal to obtain flow information, but the goal of visualizing vasculature is the same, and the resulting images are similar. At the time of this article, Carl Zeiss Meditec and Optovue have received FDA clearance for their OCTA devices. Other companies are also actively developing their OCTA systems, and some have released products outside the US.

CLINICAL APPLICATIONS

One of the greatest advantages of OCTA is its ability to provide 3-dimensional, depth-encoded information. Because it is an OCT-based scan, the data obtained consists of multiple B-scans which are combined into a cube. By segmenting the cube into different layers or slabs, OCTA is able to isolate blood flow information at a specific retinal depth. Figure 2 shows OCTA en face images of a healthy retina at various depths. The superficial capillary plexus (SCP) can clearly be seen to have a different vessel pattern than the deep capillary plexus (DCP) and choriocapillaris, consistent with histological studies7,8. This is the first imaging modality that allows noninvasive, in vivo visualization of individual capillary networks in the retina. In comparison, FA images are 2-dimensional, which means signals from all layers in the eye are projected in the same plane, thereby losing depth information.

Figure 2. A) Superficial capillary plexus. B) Deep capillary plexus. C) Choriocapillaris

 

This unique capability of OCTA gives rise to new opportunities for studying the retina and understanding the pathophysiology of diseases. For instance, paracentral acute middle maculopathy (PAMM) is a condition characterized by intermediate and deep capillary plexus infarction, and has been hypothesized to affect patients with various retinal vascular diseases such as retinal vein occlusion and sickle cell disease9. A few papers have recently demonstrated with OCTA that the SCP and DCP are affected asymmetrically in PAMM, with more ischemia observed in the DCP (Figure 3), providing angiographic evidence of this condition10,11. Segmenting OCTA data also aids in highlighting vasculature in a specific area of interest. For example, neovascularization can develop in the setting of various diseases, and generally occurs at specific retinal depths depending on the disease. In wet age-related macular degeneration, OCTA segmentation can help isolate the choroidal neovascular membrane (CNVM) at the level of the RPE (Figure 4). In diseases such as retinal vein occlusions and proliferative diabetic retinopathy, neovascularization typically occurs in the preretinal space. By segmenting above the inner limiting membrane, OCTA allows visualization of the abnormal blood vessels independently from normal retinal vessels. Figure 5 shows FA and OCTA imaging of a branch retinal vein occlusion (BRVO) with neovascularization elsewhere (NVE).  OCTA shows good agreement with FA in visualizing areas of ischemia. In addition, OCTA is able to isolate the NVE lesion by segmenting the preretinal space (Figure 5D), and the fine vessel details are not affected by dye leakage as in the FA image (Figure 5B). Isolating the lesion of interest enables clinicians to view only the information pertinent to the disease, potentially allowing a more detailed assessment of microvascular pathologies. It may also allow for easier comparison between visits to monitor for change.

Figure 3. OCTA of a patient with sickle cell maculopathy. Note that while the SCP is relatively preserved (A), the DCP shows areas of ischemia temporal to the fovea (B).

 

 

Figure 4. A) FA image of a large, mature fibrovascular scar from type 1 CNV. Because of the extensive scarring of the macula, there is hypofluorescence centrally. B) OCTA image overlaid on FA. In this case, OCTA is able to image the CNV lesion in greater detail than FA, with the mature trunks of the CNV lesion clearly visible. C) The sub-RPE space was segmented to isolate the CNV lesion.

 

Figure 5. A) Early-phase FA of a BRVO. B) Mid-phase FA. C) OCTA en face image of the retina, with corresponding B-scan. D) OCTA en face image segmenting the preretinal space, with corresponding B-scan.

 

Another advantage of OCTA is its resolution. As mentioned above, OCTA can achieve an axial resolution of approximately 5 microns, which permits high definition imaging and clear visualization of microvascular capillaries that are difficult to view with current imaging methods. In Figure 6, areas of capillary nonperfusion can be seen clearly in a patient with diabetes. Several studies have shown early but promising results in the reliability of OCTA in imaging retinal ischemia in the setting of diabetic retinopathy. In particular, small pockets of ischemia seem to be visualized better on OCTA than FA. As research is being done on preventing or reversing macular ischemia, OCTA may serve as a useful tool in monitoring retinal perfusion changes over time12. Figure 7 shows the SCP and DCP images of a patient with fovea plana. OCTA beautifully demonstrates the absence of the foveal avascular zone (FAZ) in the SCP. In the DCP, the FAZ can still be seen. Together with the structural B-scan which shows a lack of normal foveal depression, OCTA is able to easily confirm the diagnosis of fovea plana.

Figure 6. A) 3x3 micron scan and B) 6x6 micron scan of a patient with diabetic retinopathy. Areas of capillary nonperfusion can be seen as dark areas in the retina. A few microaneurysms are highlighted with yellow arrows.

 

Figure 7. OCTA images of a patient with fovea plana. A) SCP demonstrates the absence of a normal FAZ. The corresponding B-scan shows no foveal depression. B) OCTA en face image of the DCP shows that although the FAZ is absent in the SCP, it is still present in the DCP.

The noninvasive nature and short scan duration also make OCTA easy for patients to tolerate. Image acquisition usually takes a matter of seconds, compared to FA which involves a series of photographs taken over minutes. In addition, since there is no need for dye injections, OCTA may be performed more frequently than FA. This enables clinicians to follow the patient more closely, and potentially detect changes earlier. All of these features make OCTA a promising tool for diagnosing and managing disease.

LIMITATIONS / ARTIFACTS

As with any test, interpreting OCTA images requires an understanding of the limitations of the technology, as well as being able to recognize artifacts that may confound the data. Because OCTA images are derived from OCT data, they are subject to the same artifacts. Any media opacity that blocks light from reaching the retina will cause the OCTA image to appear dark. Media opacities include cataracts, floaters, and even dry eyes. In addition, dense pigment within the retina (such as hemorrhages) can obstruct light from reaching deeper layers. When there are dark areas on the OCTA en face image, sometimes it is not possible to distinguish between media opacities and true capillary nonperfusion based on the OCTA image alone. In these cases, it is important to reference the structural B-scans and structural en face image. Media opacities can be seen on the B-scan as a dark shadow, whereas capillary nonperfusion will not affect the brightness of the B-scan. In Figure 8A, there is a dark area nasal to the fovea on the OCTA en face image, which corresponds to a dark area on the structural en face image, as well as the B-scan in the corresponding location (Figures 8B, 8C). This confirms that the dark area on the angiogram is the result of a floater, not capillary nonperfusion.

Projection artifacts are the result of fluctuating shadows cast by blood flow in a more superficial layer of the retina onto deeper layers. These artifacts are detected as flow signals, and will have the same vessel pattern as the inner layer. Projection artifacts typically occur in reflective retinal layers, which appear bright on the structural B-scan, most strongly in the RPE13. When evaluating deeper retinal layers for abnormal vasculature, it is important to rule out projection artifacts to avoid false positives. Currently, OCTA software have post-processing methods to reduce or eliminate these artifacts. In Figure 9, the SCP and DCP of a healthy eye are shown (Figures 9A, 9B). The DCP slab shows some projection artifacts from the SCP, which appear as vessels that replicate the pattern of the SCP. The same slab with projection artifacts removed is shown in Figure 9C. Figure 10 shows a subretinal, type 2 CNV. Because the CNV lesion is close to the level of the RPE, projection artifacts from both the SCP and DCP can be seen (Figure 10A). After removing the artifacts, the CNV lesion is clearly isolated.

Figure 8. A) OCTA en face image with a dark area caused by a floater (yellow dashed circle). B) The structural en face image shows the same artifact in the same location. C) B-scan through the area shows are dark shadow through all layers of the retina, suggesting that this is a vitreal floater.

 

Figure 9. A) SCP of a healthy eye. B) DCP without projection artifact removal. C) DCP with projection artifact removal.

 

Figure 10. A) Type 2 CNV close to the level of the RPE, without projection artifact removal. B) Same lesion with projection artifact removal.

 

Areas with slow flow are also more difficult to image with OCTA. Due to the fast scanning speed of OCT, areas with slower flow may not have enough difference between successive B-scans to meet the OCTA detection threshold. Therefore, any flow that is slower than the detection threshold will not be shown on OCTA images, and will appear dark. Examples of areas with slow flow include intraretinal fluid, some microaneurysms, and hemorrhages. However, many of these areas can be visualized on the B-scan, such as in Figure 11.

Figure 11. A) OCTA en face image of the retina of a patient with diabetic macular edema. Retinal ischemia around the FAZ can be seen, but intraretinal fluid and leakage are not visible. B) B-scan shows intraretinal fluid at the macula.

Finally, segmentation errors can be a source of artifacts. The appearance of the OCTA en face image is dictated by its corresponding slab, whose upper and lower boundaries are defined by segmentation lines. Therefore, any errors in segmentation will cause the en face angiogram to be displayed incorrectly. Figure 12 is an example of a patient with multifocal choroiditis. In Figure 12A, there appears to be an area within the inflammatory lesion that looks suspicious for choroidal neovascularization. However, a closer look at the B-scan reveals an error in segmentation in that same area. After the segmentation is corrected, the bright area on the angiogram is no longer present (Figure 12B). Checking segmentation is important when evaluating OCTA images, and gives the clinician more confidence in the angiogram’s findings.

 

Figure 12. A) Choriocapillaris with segmentation error. B) Choriocapillaris without segmentation error.

 

 

OCTA vs. FA

Because they both have the word “angiography” as part of their names, it is natural for clinicians to draw comparisons between the two imaging modalities, and even explore the possibility of replacing FA with OCTA. It is important to note that while both OCTA and FA aim to image blood flow, they are different technologies, and will therefore image vascular features differently. Although OCTA generally shows good agreement with FA, it does not, at least presently, serve as a replacement for FA.

As mentioned above, OCTA is not able to image leakage. In addition, staining, pooling, and vessel filling time are also not features of OCTA. However, this also means that vasculature below an area of fluid, which would otherwise be obscured by dye leakage with FA, can be seen on OCTA4. Since they each have their strengths and weaknesses, determining which test is indicated for the patient should be based on clinical need. Most likely, OCTA will serve as a complementary test to FA, giving clinicians additional information for assessing retinal vasculature.

CONCLUSION

Although OCTA is a technology that is still in its infancy, it has already shown to have great potential for the future. The ability to provide depth-encoded information at very high resolution makes it a valuable tool for both clinicians and researchers. Its noninvasive nature also makes it readily accessible to optometrists to utilize.

Future developments in OCTA technology will further optimize its clinical utility. Automatic quantification of vessel density may be useful in monitoring diseases such as diabetes and glaucoma14,15; wider fields of view and improved resolution will allow clinicians to utilize OCTA in more situations; there is also much research currently ongoing to explore the potential new applications of this imaging method. OCTA is a promising new technology which is generating great interest, and is likely to become more popular in clinical practice over the next few years.

REFERENCES

  1. Bennett, Timothy J., et al. "Fundamentals of fluorescein angiography." Curr Concepts Ophthalmology 9.3 (2001): 43-49.
  2. de Carlo, Talisa E., et al. "A review of optical coherence tomography angiography (OCTA)." International Journal of Retina and Vitreous 1.1 (2015): 1.
  3. Choi, WooJhon, et al. "Choriocapillaris and choroidal microvasculature imaging with ultrahigh speed OCT angiography." PloS one 8.12 (2013): e81499.
  4. Puliafito Carmen, et al. OCT Angiography: The Next ‘Big Thing’ in Retinal Imaging. Vindico Medical education. 2015 April.
  5. Zhang, Anqi, et al. "Methods and algorithms for optical coherence tomography-based angiography: a review and comparison." Journal of biomedical optics 20.10 (2015): 100901-100901.
  6. Kogachi, K, and A. Kashani. "Optical Coherence Tomography Angiography: A Primer on This Promising New Technology for Evaluating Retinal Vasculature." New Retina MD 2 (2015): 6-8.
  7. Henkind, P. "Radial peripapillary capillaries of the retina. I. Anatomy: human and comparative." The British journal of ophthalmology 51.2 (1967): 115.
  8. Matsunaga, Douglas, et al. "OCT angiography in healthy human subjects." Ophthalmic Surgery, Lasers and Imaging Retina 45.6 (2014): 510-515.
  9. Minvielle, Wilfried, et al. "Macular microangiopathy in sickle cell disease using optical coherence tomography angiography." American journal of ophthalmology (2015).
  10. Ilginis, Tomas, Pearse A. Keane, and Adnan Tufail. "Paracentral acute middle maculopathy in sickle cell disease." JAMA ophthalmology 133.5 (2015): 614-616.
  11. Nemiroff, Julia, et al. "Assessing Deep Retinal Capillary Ischemia in Paracentral Acute Middle Maculopathy by Optical Coherence Tomography Angiography." American journal of ophthalmology 162 (2016): 121-132.
  12. Bradley, Patrick D., et al. "The Evaluation of Diabetic Macular Ischemia Using Optical Coherence Tomography AngiographyOCTA in Diabetic Retinopathy." Investigative ophthalmology & visual science 57.2 (2016): 626-631.
  13. Spaide, Richard F., James G. Fujimoto, and Nadia K. Waheed. "Image artifacts in optical coherence tomography angiography." Retina 35.11 (2015): 2163-2180.
  14. Jia, Yali, et al. "Quantitative OCT angiography of optic nerve head blood flow." Biomedical optics express 3.12 (2012): 3127-3137.
  15. Hwang, Thomas S., et al. "Automated Quantification of Capillary Nonperfusion Using Optical Coherence Tomography Angiography in Diabetic Retinopathy." JAMA ophthalmology (2016): 1-7.

                                            Click Here to Take the Test

COA Champion Supporters:

California Optometric Association
2701 Del Paso Road, Ste. 130-398 | Sacramento, CA 95835 | 833-206-0598