Future Directions in Retinal Optical Coherence Tomography
A review of new technologies on the horizon.
Veena Raiji, MD • Alexander Walsh, MD • Srinivas Sadda, MD
|Veena Raiji, MD, Alexander Walsh, MD, and Srinivas Sadda, MD, are on the faculty of the Doheny Eye Institute at the Keck School of Medicine of the University of Southern California. Dr. Sadda is a consultant to Heidelberg Engineering, receives research support from Carl Zeiss Meditec and Optovue, and shares in royalties from intellectual property licensed to Topcon Corporation by the Doheny Eye Institute. Dr. Walsh shares in royalties from software licensed to Topcon Corporation by the Doheny Eye Institute. Dr. Raiji reports no financial interest in any products mentioned here. Dr. Sadda can be reached via e-mail at SSadda@doheny.org.|
First reported by Huang et al. in 1991, optical coherence tomography represents a noninvasive method for cross-sectional imaging of the internal retinal structures through the detection of optical reflections and echo time delays of light, using low-coherence interferometry to subsequently produce in vivo two-dimensional images of internal tissue microstructure.1,2
Similar to ultrasound, however using light instead of sound, OCT was first demonstrated to provide both quantitative and qualitative assessment in the peripapillary area of the retina, as well as the coronary artery, and it can detect reflected signals as small as 10-10 of the incident optical power with longitudinal and lateral spatial resolutions of a few microns.1
Interferometry measures the difference between reflected light waves from a tissue of interest and that from a reference path using low-coherence light.3 Interference is detected when the difference in path length between these two is less than the coherence length of the light source.2 Light reflected from the retinal structures interferes with the reference beam light, and the interference of echoes is detected, creating a measurement of light echoes vs depth.3,4
Long-coherence light, such as that from superluminescent diodes or femtosecond lasers, typically produces images with poor axial resolution, whereas with low-coherence light, interference occurs over short distances, and high axial resolution is possible.4 Axial resolution is determined by the bandwidth of the light source and the coherence grading (dependent upon the central wavelength).4
Transverse resolution is limited by confocal gating, or the size of the light spot that can be focused on the retina; however, the best structural resolution occurs only when light is focused on by the tissue of interest.7 Additionally, the absorption of central light by the tissue of interest must be considered.
When light sources between 200 and 600 nm are used, hemoglobin absorbs a significant proportion of the incident light.7 To balance these properties, a central wavelength with high axial resolution and low scattering is ideal. Deeper penetration is limited by light scatter from optical media, as well as melanin light absorption at the level of the retinal pigment epithelium.4
Since its advent in 1991, OCT has become a main-stay in retinal imaging and, in some situations, has supplanted fluorescein angiography. Initial devices utilized time-domain (TD) technology (Stratus OCT, Carl Zeiss Meditec), which employed a mobile reference arm mirror that sequentially measured light echoes from time delays with acquisition speeds of 400 A-scans/second and axial resolution of 8-10 µm.2,6
Time-domain technology was limited by long acquisition times (due to a mobile reference mirror) and subsequent limited image sampling, which potentially overlooked small macular lesions. Additionally, motion artifacts and patient blinking limited TD-OCT's resolution.8,9
Fourier domain (FD) schemes succeeded TD technology, and they use a central wavelength of 800-850 nm, a stationary reference arm, a high-speed spectrometer, and a charge-coupled device (CCD) line-scan camera. The reference mirror does not require mechanical scanning to detect light echoes simultaneously, which increased acquisition speed to 25,000-52,000 A-scans/second, with axial resolution of 3-7 µm, significantly improving the signal-to-noise ratio and allowing detection of individual retinal layers and lesion components, which is particularly useful in disease applications, such as neovascular AMD.6,10
Spectral-domain OCT represents a significant improvement over TD-OCT in axial image resolution and image acquisition, reduction of motion artifacts, increased area of retinal coverage, and the ability to produce three-dimensional data sets to create topographic maps with precise registration.2 Internal limitations of commercially available CCD line-scan cameras limit the transfer of interferometric signals, which in turn limits axial resolution and acquisition speed.4,9
Additionally, SD-OCT is still subject to some element of motion artifacts, segmentation artifacts, and interinstrument comparability.6,9 Known limitations of both TD- and SD-OCT, which several investigators are trying to overcome, include limited resolution due to infrared radiation absorption by anterior-segment structures and ocular media, limited axial resolution due to image scattering from ocular structures (so-called speckle noise), and limited lateral resolution due to the restricted numerical aperture of the optical system.10
Another form of FD-OCT, swept source (SS)-OCT technology uses a narrowband light source with central wavelength of ~1,050 nm, with a short cavity-swept laser (instead of a superluminescent diode laser) that can emit different frequencies of light and that is rapidly tuned over a broad bandwidth.2,5,9 It uses a high-speed complementary metal oxide semiconductor (CMOS) camera and two parallel photodetectors to achieve 100,000-400,000 A-scan/second rates with 5.3-µm tissue axial resolution over a 4-mm imaging range.
The small-area, high-density imaging (with extensive B-scan averaging to reduce speckle noise artifact) allows imaging down to the level of individual photoreceptors, particularly when coupled with adaptive optics (see below).9 Compared to SD-OCT, SS-OCT shows reduced fringe washout (reduced signals at the edges of the B-scans), better sensitivity with imaging depth, longer imaging range, higher detection efficiencies, and the ability to perform dual balanced detection.
These advantages allow less patient-introduced artifacts from movement and breathing, as well as better penetration through cataracts and ocular opacities. The long imaging range (~7.5 mm) also allows the anterior segment to be evaluated without directly employing complex imaging techniques, typically subject to artifacts.9
Despite increased acquisition speed and subsequently reduced motion artifacts, all in vivo imaging is still subject to some blur artifacts and image distortion, even under the best patient cooperation circumstances.11 Improvements in dynamic retinal tracking and registration algorithms have helped to eliminate some of these artifacts to allow improved resolution.
Current OCT technology has a shallow penetration depth (~1-3 mm), limiting its potential applications in the sub-RPE space and choroid.4,12 Image formation in OCT depends on photon scattering — some photons are “singly” scattered, while others are scattered “multiply.” Those that are singly scattered add to the OCT signal, while those that are scattered multiple times contribute to background noise.12
Additionally, the large water composition of the eye limits the light wavelengths that can be used.13 The absorption spectrum of water, in regions that are still safe for optical exposure, has two regions in which light absorption is low, one up to ~950 nm and the other between 1,000 and 1,100 nm.14
The proportions of scattered photons, the absorption spectrum of water, the potential for light scatter (optical media), and absorption (melanin) all dictate the potential depth of choroidal penetration. Previously difficult to utilize due to cumbersome operating systems and camera limitations, devices utilizing the 1,000-1,100-nm wavelength may safely allow enhanced sub-RPE imaging with axial resolution in the 8-µm range, still with ultrahigh-speed image acquisition.4
Sub-RPE imaging is specifically useful in the management of suspected choroidal disorders, such as central serous chorioretinopathy, AMD, choroidal tumors and retinitis pigmentosa (Figures 1 and 2).4
Figure 1. Choroidal melanoma 1.7 mm in height.
Figure 2. Even when the choroid was fully visible at 840 nm (left), considerable additional detail was visible at 1,050 nm (right).
Adaptive optics has improved upon already existing OCT techniques by correcting for higher-order ocular aberrations during image acquisition, allowing near cellular level resolution.15,16 Ultrahigh-resolution adaptive optics OCT (AOOCT) uses a high-speed CMOS camera with a novel image-registration/dewarping algorithm to limit motion artifacts, increase lateral resolution, reduce speckle and enhance sensitivity.16
With a pupil diameter of >6 mm, lateral resolution of 2-3 mm can be achieved, which is sufficient for the resolution of individual cones on a three-dimensional basis, even in the same 0.5º retinal patch, and of crosssectional profiles of individual nerve fiber layer bundles and foveal capillaries and to define the foveal avascular zone.11,16 In combination with an adaptive-optics slit-lamp ophthalmoscope, AO-OCT can acquire both OCT and SLO in vivo with ≤3.5-µm resolution.15
FULL FIELD OCT
Full-field (FF)-OCT is based on spatial coherence grating and uses narrow-band illumination with high-numerical aperture objectives and a liquid crystal retarder to minimize defocusing and dispersion effects, producing three-dimensional imaging with ultrahigh resolution. Safrani et al. successfully demonstrated the nucleus of an onion cell using FF-OCT.17
Our ability to detect clinically significant disease requiring surgical intervention (macular holes, epiretinal membranes, retinal detachments, traction) has been revolutionized by the introduction of OCT. The use of intraoperative SDOCT augments traditional intraoperative microscopy to help surgeons better delineate tissue structures, reducing surgical times and excessive illumination and limiting the need for potentially toxic stains.18
Implementing a superluminescent diode with a center wavelength of 840 nm and bandwidth of 49 nm, microscope-mounted (MM)-OCT can scan a 12-mm field of view, allowing real-time crosssectional analysis of tissue structures, the shape or position of which may change intraoperatively.
Microscope-mounted OCT was designed to work with the optical path of the Oculus BIOM3 suspended from a Leica M841 ophthalmic surgical microscope, and it provides a high-magnification telescope with a large field of view to the view port of the surgical microscope.18 Metallic surgical instruments are detected as highly reflective with total shadowing below the instrument, while polyamide materials are moderately reflective with subtotal shadowing, and silicone instruments are moderately reflective with minimal shadowing.19
Microscope-mounted OCT demonstrated interactions between surgical instruments and retinal tissue in cadaveric porcine eyes.19 While MM-OCT is advantageous compared to handheld OCT because it is stabilized and utilizes a common focal plane to the microscope optical path, it is still limited by its 840-nm center wavelength and lateral resolution.18
Widefield OCT employs swept-source technology to evaluate larger portions of the central retina. Ultrahigh-speed SS-OCT, using a 1,050-1,060-nm Fourier-domain mode locked laser, collects 1,900 x 1,900 A-scans with a roughly 70º angle of view (~1 cm2) in 3-6 seconds with image acquisition speeds between 684,000 and 1,368,700 Ascans/second. These datasets are then consolidated into a 4 megapixel high-definition image.
Additionally, good choroid and choroidal/scleral interface penetration can be achieved with axial resolution of 6.7-19 µm.20,21 In combination with optical microangiography (OMAG) technology, vascular perfusion mapping, down to the capillary level, is possible. OMAG technology utilizes a 840-nm wavelength with an A-scan rate of 27,000 Hz and an axial resolution of 8 µm to image a 7.4 x 7.4 mm2 area of the posterior segment, allowing volumetric map acquisition, comparable to fluorescein and indocyanine green angiography.22
By measuring the Doppler shift and relative angle between the OCT beam and a blood vessel, blood flow velocity can be assessed.2,23 In patients with perimetric glaucoma, decreased retinal blood flow assessed by Doppler OCT (840 nm wavelength, axial resolution of 5 µm, transverse resolution of 20 µm) was evident, even in the absence of structural loss, assessed by retinal nerve fiber layer (rNFL) OCT.24
Using this technology, blood-flow measurement can be assessed from the transection of all branch retinal arteries and veins by eight circular scans, acquired in ~2 seconds, each composed of 3,000 axial scans and then summing all blood flow measured in the veins.25 The background axial inner retinal tissue boundary motion is compared with that of the vessel wall to attain the net Doppler shift, produced by blood flow.25
In normal subjects, the mean average total retinal blood flow is estimated to be 45.6 mL/min.25 Such technology will have applications in any disease with a vascular component, such as proliferative diabetic retinopathy, ischemic optic neuropathy, and glaucoma.23
Polarization sensitive (PS)-OCT is a method of functional tissue assessment by means of light polarization evaluation.26 PS-OCT allows individual retinal layer identification by measuring cross-sectional and volumetric birefringence, contrasting between birefringent layers and other retinal layers.2,27 PS-OCT simultaneously measures intensity (conventional OCT images), retardation, and optic axis orientation to distinguish polarization-preserving tissue, birefringent tissue, and polarization-scrambling tissue.26
Birefringent tissues include the rNFL, Henle's layer, sclera, or any fibrotic tissue that increase phase retardation. The RPE is a polarization-scrambling layer. Light transmitted through the RPE maintains the same polarization state and degree of retardation as images below and above the RPE layer, allowing assessment of RPE damage, which is specifically useful in the context of pigment epithelial detachments and pseudovitelliform dystrophy.26
Focal polarization scrambling within the neurosensory retina may be an indication of retinal thickness, and unspecified RPE thickening may represent fibrosis — an important therapeutic target in the management of several disease processes.26 Yamanari et al. described a PSOCT–utilizing swept-source technology for 1-µm wavelength imaging to visualize the sclera and lamina cribrosa in vivo with high sensitivity, based on similar birefringence patterns.27
In summary, a variety of emerging OCT technologies are poised to expand significantly the scope of OCT imaging and to enhance significantly our approach to the diagnosis and management of patients with retinal disease. An in-depth understanding of these technologies and their potential advantages and disadvantages will aid retinal specialists in optimally using these new methods. RP
1. Huang D, Swanson EA, Lin CP, et al. Optical coherence tomography. Science. 1991;254:1178-1181.
2. Duker JS. OCT technology: what is coming? Paper presented at: Annual meeting of the European Society of Retina Specialists; Rome, Italy; January 28, 2012.
3. Costa RA, Skaf M, Melo LA Jr, et al. Retinal assessment using optical coherence tomography. Prog Retin Eye Res. 2006;25:325-353.
4. Keane PA, Ruiz-Garcia H, Sadda SR. Clinical applications of long-wavelength (1,000-nm) optical coherence tomography. Ophthalmic Surg Lasers Imaging. 2011;42:S67-S74.
5. Regatieri CV, Branchini L, Duker JS. The role of spectral-comain OCT in the diagnosis and management of neovascular age-related macular degeneration. Ophthalmic Surg Lasers Imaging. 2011;42:S56-S66.
6. Sull AC, Vuong LN, Price LL, et al. Comparison of spectral/Fourier domain optical coherence tomography instruments for assessment of normal macular thickness. Retina. 2010;30:235-245.
7. van Velthoven MEJ, Faber DJ, Verbraak FD, van Leeuwen TG, de Smet MD. Recent developments in optical coherence tomography for imaging the retina. Prog Retin Eye Res. 2007;26:57-77.
8. Keane PA, Sadda SR. Imaging chorioretinal vascular disease. Eye. 2009;24:422-427.
9. Potsaid B, Baumann B, Huang D, et al. Ultrahigh speed 1050nm swept source/Fourier domain OCT retinal and anterior segment imaging at 100,000 to 400,000 axial scans per second. Opt Express. 2010;18:20029-20048.
10. Wojtkowski M, Kaluzny B, Zawadzi RJ. New directions in ophthalmic optical coherence tomography. Optom Vis Sci. 2012 Mar 22. [Epub ahead of print].
11. Kocaoglu OP, Lee S, Jonnal RS, et al. Imaging cone photoreceptors in three dimensions and in time using ultrahigh resolution optical coherence tomography with adaptive optics. Biomed Opt Express. 2011;2:748-763.
12. Sharma U, Chang EW, Yun SH. Long-wavelength optical coherence tomography at 1.7 microm for enhanced imaging depth. Opt Express. 2008;16:19712-19723.
13. Unterhuber A, Povazay B, Hermann B, Sattmann H, Chavez-Pirson A, Drexler W. In vivo retinal optical coherence tomography at 1040 nm: enhanced penetration into the choroid. Opt Express. 2005;13:3252-3258.
14. American National Standards Institute. American National Standard for Safe Use of Lasers. New York, NY: Author; 2000.
15. Zawadzki RJ, Jones SM, Pilli S, et al. Integrated adaptive optics optical coherence tomography and adaptive optics scanning laser ophthalmoscope system for simultaneous cellular resolution in vivo retinal imaging. Biomed Opt Express. 2011;2:1674-1686.
16. Miller DT, Kocaoglu OP, Wang Q, Lee S. Adaptive optics and the eye (super resolution OCT). Eye (Lond). 2011;25:321-330.
17. Safrani A, Abdulhalim I. Ultrahigh-resolution full-field optical coherence tomography using spatial coherence gating and quasi-monochromatic illumination. Opt Lett. 2012;37:458-460.
18. Tao YK, Ehlers JP, Toth CA, Izatt JA. Intraoperative spectral domain optical coherence tomography for vitreoretinal surgery. Opt Lett. 2010;35:3315-3317.
19. Ehlers JP, Tao YK, Farsiu S, Maldonado R, Izatt JA, Toth CA. Integration of a spectral domain optical coherence tomography system into a surgical microscope for intraoperative imaging. Invest Ophthalmol Vis Sci. 2011;52:3153-3159.
20. Klein T, Wieser W, Eigenwillig CM, Biedermann BR, Huber R. Megahertz OCT for ultrawide-field retinal imaging with a 1050 nm Fourier domain modelocked laser. Opt Express. 2011;19:3044-3046.
21. Povazay B, Hermann B, Hofer B, et al. Wide-field optical coherence tomography of the choroid in vivo. Invest Ophthalmol Vis Sci. 2009;50:1856-1863.
22. An L, Subhush HM, Wilson DJ, Wang RK. High-resolution wide-field imaging of retinal and choroidal blood perfusion with optical microangiography. J Biomed Opt. 2010;15:026011.
23. Sadda S. Future OCT technology. Paper presented at: Macula 2012; New York, NY; January 21, 2012.
24. Hwang JC, Konduru R, Zhang X, et al. Relationship among visual field, blood flow, and neural structure measurements in glaucoma. Invest Ophthalmol Vis Sci. 2012 Mar 23. [Epub ahead of print]
25. Wang Y, Lu A, Gil-Flamer J, Tan O, Izatt JA, Huang D. Measurement of total blood flow in the normal human retina using Doppler Fourier-domain optical coherence tomography. Br J Ophthalmol. 2009;93:634-637.
26. Michels S, Pircher M, Geitzenauer W, et al. Value of polarisation-sensitive optical coherence tomography in diseases affecting the retinal pigment epithelium. Br J Ophthalmol. 2008;92:204-209.
27. Yamanari M, Lim Y, Makita S, Yasuno Y. Visualization of phase retardation of deep posterior eye by polarization-sensitive swept-source optical coherence tomography with 1-micron probe. Opt Express. 2009;17:12385-12396.