Simultaneous OCT/SLO/ICG Imaging

Abstract

purpose. To evaluate how information from combined coronal optical coherence tomography (OCT) and confocal laser scanning ophthalmoscopy (SLO) with integrated simultaneous indocyanine green (ICG) dye angiography can be used in the diagnosis of a variety of macular diseases.

methods. A compact chin-rest-based OCT/confocal imaging system was used to produce the OCT image and excite the fluorescence in the ICG dye. The same eye fundus area can be visualized with coronal (C-scans, en face) OCT and ICG angiography simultaneously. Fast T scanning (transverse scanning, en face) was used to build B- or C-scan OCT images along with confocal SLO views, with and without ICG filtration. The OCT, confocal SLO and ICG fluorescence images were simultaneously presented in a three-screen format. A live mixing channel overlaid the ICG sequence on the coronal OCT slices in a fourth panel for immediate comparison.

results. Thirty eyes were imaged. The pathologic conditions studied included classic and occult neovascular membranes, vascularized RPE detachments, polypoidal choroidal vasculopathy, traumatic choroidal rupture, diabetic maculopathy, central serous retinopathy, and macular drusen. Images were evaluated with special attention toward identifying novel relationships between morphology and function revealed by the superimposition of the studies.

conclusions. Simultaneous visualization of an en face (coronal, C-scan) OCT image and of an ICG angiogram, displayed side by side and superimposed, permits more precise correlations between late fluorescence accumulation with structures deep to the retinal surface at the retina-choroid interface. The multiplanar scanning also permits immediate B-scan OCT cross-sectional views of regions of abnormal fluorescence. The paper demonstrates the synergy between the two types of studies, functional and anatomic, in providing a more complete view of the pathologic condition.

Fluorescence imaging of indocyanine green (ICG) dye ^{1 2} is an established technique used in ophthalmic investigations of the choroidal circulation of the eye. ICG is a well-tolerated drug, and the infrared light has a higher threshold for causing phototoxic effects in the retina than shorter wavelength light. ³  

The resolution in fluorescence imaging is usually limited by the imaging instrument used, such as the confocal microscope used in this study. In the case of retinal imaging, a resolution on the order of 300 μm represents the maximum achievable with confocal scanning laser ophthalmoscopy (SLO). Therefore, the depth accuracy in the evaluation of microvolumes of leakage is in the range of 0.5 mm. Higher depth resolution is achievable in imaging the retina by using principles of optical coherence tomography (OCT). ⁴ However, OCT relies on interference, where the signal from the retina interferes with that in the reference arm of the interferometer, where both signals are generated by the same optical source. This principle is not applicable to fluorescence imaging. Therefore, improved compounding of information could be achieved only by the development of a combination of these two principles: confocal (fluorescence) and OCT imaging. 

The utility of combining information from these two techniques has been described by previous investigators. ⁵ The combination of techniques may prove more useful in those circumstances ⁶ in which ICG alone is not sufficient for accurate diagnostic or where it does not add anything substantial to the investigation. 

There are also often cases in which separate instruments are used—an ICG angiogram and a series of OCT images ⁷ —to help attain more data about the fundus and correlate the features observed. The angiographic images display coronal orientation (C-scans), although OCT images from conventional systems are sagittal in orientation (B-scans). The method described in this paper and the system we assembled provide C-scan images in both channels, ICG and OCT, as well as conventional B-scan OCT images. This facilitates the comparison of features between the two modalities and enhances the localization of small lesions within the angiogram, as well as offering better guidance of the B-scan orientation and position based on the information in the two C-scan images, ICG, or OCT. 

The technology used here is based on assembling OCT images from T-scans. T-scans are transversely acquired, one-dimensional reflectivity profiles. This technology is different from that based on spectral-domain OCT (SD-OCT), in which images are formed from A-scans (axial reflectivity profiles). SD-OCT is rapidly becoming more favored in ophthalmology ⁸ than the time-domain OCT (TD-OCT), because of its speed and sensitivity advantage. ⁹ Several companies have commercialized stand-alone units, and a system from Heidelberg Engineering (Heidelberg, Germany) ¹⁰ combining SD-OCT with angiography recently became commercially available. Other attempts in using SD-OCT evolved in a no-dye angiography method, ¹¹ in which the vessel structure is separated from the SD-OCT signal to generate an equivalent angiography image. 

The SD-OCT method, although generating much faster cross-sectional (B-scan) images than TD-OCT systems (such as the en face OCT presented herein), still requires significantly longer acquisition and processing time to generate an en face (coronal or C-scan) image. ¹² This process requires the collection of the whole volume of the sample, processing the data to produce a 3-D volume and then off line interpolation of the data volume to select a C-scan-oriented cut. Even the fastest SD-OCT research systems ¹³ require at least a second or more. To achieve 3-D reconstructions in a matter of seconds, transverse resolution is always compromised, which leads to lower transverse resolution in the en face software reconstructed image. 

The en face OCT technology used in our study offers the examiner the possibility of rapidly acquiring sequential B- and C-scans by switching the instrument between the two regimens. The en face scans provide a simultaneous side-by-side comparison with the familiar appearance of the fundus as seen in a direct ophthalmoscope or in an SLO system. Sequential and rapid switching between the en face viewpoint and the cross-sectional viewpoint, specific for the en face OCT systems, represents a significant advantage in the acquisition, as images with different orientations can be obtained using the same system. 

This flexibility for obtaining images in orthogonal planes in real time is an advantage that is not available in even the fastest spectral system. In the current implementation of our system, the frame rate is 2 Hz (i.e., a C-scan is produced in real time in 0.5 seconds alongside the confocal image, which is less than any value reported by using SD-OCT to infer a C-scan by means of computer software). ¹³ As a supplementary advantage, in our system the transverse resolution is the same in both B- and C-scans, as are the acquisition rates, and the switching between one regimen and the other requires only the pressing of a keyboard key. 

Methods

We previously reported a combined OCT/ICG fluorescence instrument ¹⁴ which provided two en face (C-scans) images at the same time: an angiography image and an OCT image. Even though fluorescence leakage was still displayed with the resolution of the cSLO channel, it was possible to evaluate the morphology around the leakage with much better depth resolution in the OCT channel. The instrument opened several avenues of research in how to best combine the two images to expand the diagnostic power of the technique. In that version of the system, the confocal channel was used for either imaging at the excitation wavelength, as in an SLO system, or imaging the fluorescence distribution by introducing a band-pass filter in front of the detector of the SLO system, an avalanche photodetector (APD). That system had the disadvantage of losing the guidance capability of the confocal channel in the first seconds after the filter was introduced in the front of the detector and until the ICG reached the eye. In those circumstances the confocal image is dark, and the user cannot address the eye or head movements. In addition, there may be value to displaying the SLO image with its better morphology characterization alongside the ICG image which displays only the distribution of the fluorescent dye. 

The system presented herein features a third channel that enables the two confocal images to be displayed at the same time. The simultaneous display of three screens provides a wealth of information to the clinician. In this article we report on the use of the information acquired with these three imaging modalities and evaluate how features observable in one image can be used in the interpretation of features in the other images. 

The instrument, presented in Figure 1 , is a triple-channel OCT/confocal ophthalmoscope/ICG angiograph with versatile scanning and image display capabilities that allow the acquisition of OCT and dual confocal images in a B- or C-scan regimen. ¹⁵ The splitter used in the OCT/ophthalmoscope configuration ¹⁶ to divert some of the light to a separate confocal receiver was replaced in the present implementation by a chromatic splitter, CS1. This device separates the retina-scattered light at the excitation wavelength of 792 nm, guided into the OCT channel, from the fluorescence signal, centered at 830 nm, guided toward the fluorescence confocal receiver. To minimize the distortion of the OCT depth sampling profile (determined by the correlation function of the source), CS1 is used in transmission by the ICG and SLO channels and in reflection by the OCT channel. The residual transmission of the chromatic splitter CS1 (approx. 4%) at the OCT wavelength is sufficient to generate an SLO image. The fluorescence signal centered at 830 nm and the residual signal at the excitation wavelength 792 nm are separated by another chromatic splitter, CS2. To enhance the contrast of fluorescence in the confocal receiver, a supplementary fluorescence emission filter is used in the fluorescence channel to attenuate any excitation band light that gets past CS2. 

CS1 and CS2 are cold mirrors with a transition wavelength, λ_tr, between the excitation band and fluorescence band. Superlum (Moscow, Russia) developed a comparatively powerful superluminescent diode (SLD) for this project, with an output power of 5 mW ex fiber at λ = 792 nm and Δλ = 21 nm spectral FWHM, which determines a depth resolution in the tissue in the OCT channel of less than 9 μm (considering an index of refraction, n ≅ 1.4). 

The scanning procedure is similar to that used in any SLO system, where the fast scanning is en face (line rate, using the scanning mirror MX) and the frame scanning is much slower (at the frame rate, using the scanning mirror MY). ¹⁷ The MX mirror is driven with a ramp at 500 Hz, whereas the MY mirror is driven with a ramp at 2 Hz. In this way, an en face image in the plane (x, y) is generated at constant depth. The next en face image at a new depth is then generated by altering the length of the reference path of the OCT interferometer, by controlling a translation stage in the “reference path adjustment” block and repeating the (x, y) scan. 

To construct B-scan images, the MY mirror is kept stationary. The line scanner is driven with the same signal as in the C-scanning regimen, and the reference path adjustment block modifies the length of the interferometer arm continuously over the designated depth range in 0.5 seconds In this case, an OCT cross-sectional image is produced either in the plane (x, z) or (y, z). 

Current systems in ophthalmology now use spectral domain methods, based on either the Fourier domain OCT (using a line CCD camera) or swept-source (SS)-OCT (using a tuneable laser). Both are A-scan-based systems, and therefore the output image is a cross section (a B-scan). C-scans can be generated by postprocessing software from a stack of B-scan images collected at different lateral positions from the retinal volume. 

In our system, both C- and B-scan images are hardware obtained, by toggling a keyboard key. This method permits firing sequences of en face imaging (producing C-scans) and cross-sectional imaging (B-scans). The feature is not achievable in even the fastest research systems reported today using SS-OCT, ¹³ because the collection of a whole volume still requires 1 second, which is then followed by software slicing (i.e., to output a C-scan image takes more than 1 second). 

In the current commercially available systems that combine SD-OCT with fundus imaging (including fluorescence imaging) the two technologies used are different: A-scan- and T-scan-based, respectively. This variation may result in some offset or errors in the pointing of scanning beams from one regimen to the other. 

By contrast, our system operates in both regimens in T-scan mode (i.e., T-scans are used to assemble both C- and B-scan images). This method simplifies the software requirements and hardware demands of ensuring pixel-to-pixel correspondence. 

Images in orthogonal planes can be obtained in firing sequences. In the current implementation, the frame rate is 2 Hz (i.e., a C-scan is produced in 0.5 seconds, in real time). The system could also operate at 4 Hz, with some reduction in the transverse resolution; however, all images presented were obtained at a frame rate of 2 Hz. The transverse resolution determined by the number of pixels in the T-scans which are used to assemble B- and C-scans is the same in both scans. 

After the injection of ICG solution into the patient’s bloodstream, light from the SLD, guided through to the posterior pole of the eye by means of the interface optics, generates on one hand a reflected/backscattered return in a 21-nm band centered around 792 nm, which coherently combines with reference light to produce the OCT images, and on the other hand serves to excite fluorescence in any retinal or choroidal structures containing the ICG dye. The acquisition of fluorescent images must proceed rapidly, in less than a minute, because of the fast ICG disappearance rate from the blood stream of between 18% and 24% per minute. Generating OCT, SLO, and ICG images at a 2-Hz frame rate was found to be a reasonably good tradeoff between the acquisition speed requirement and the quality of the OCT images in terms of their signal-to-noise ratio. No eye-tracking or bite bars were used. Images with large movement artifacts can be ignored or discarded and data acquisition repeated. Also, in the images presented herein, no alignment was performed. The images shown are those obtained from the frame grabber with no other processing. 

The image in the OCT channel is generated using the path modulation created by the transverse resolution beam scan across the target. ¹⁸ For lateral image size over 1 mm, there is no need for an external phase modulator. ¹⁹  

Results

The research adhered to the tenets of the Declaration of Helsinki, and informed consent was obtained after the nature and possible consequences of the study were explained. Subjects’ ages ranged from 22 to 91 years. 

The 12-bit grayscale images in the three channels are displayed simultaneously. Two formats were possible: either side-by-side, three on a screen, or alternatively along with a fourth channel which displays the real-time overlay of the ICG on the C-scan OCT. 

We reviewed the first 30 patients, who had a variety of diseases, imaged with the system. Table 1summarizes these patients (age, disease) and the respective key findings in each of the channels (OCT, SLO, ICG, and overlay of OCT and ICG), along with some brief interpretive comments. 

In 30 seconds, 60 such groups of three images are acquired while the depth is explored over a range of typically 1.3 mm in retinal tissue. If the eye moves considerably during the acquisition and essential parts from the retina volume are missing, the acquisition can be repeated after the ICG bolus has passed. The pixel-to-pixel correspondence between the three simultaneously acquired images allows later comparison of morphologic features seen in the different imaging modes. Generally, just a few good-quality en face OCT images from the stack collected during the bolus passage are sufficient for evaluation of each phase of an imaging study. 

The first case we present is that of a patient with choroidal neovascular membrane (Table 1 , patient 10) who had been imaged before with ICG ²⁰ and OCT (Khan J. IOVS 2004;45:ARVO E-Abstract 2979). ²¹ The example shown in Figure 2demonstrates how definition of the underlying feeder vessel and choroidal neovascular membrane in a poorly defined macular lesion was achieve through the combination of imaging approaches. 

In Figure 2 , the top row is a color fundus photograph showing associated hemorrhages in the fundus. The following three rows display the three-screen format: The first column is the ICG, the second is the C-scan OCT, and the third is the SLO. The SLO images demonstrate a low-lying serous elevation of the macula. In the second row, the ICG frame on the left shows the earliest part of the vascular filling of the choroid revealing a central diagonal feeder vessel of a bicycle-shaped neovascular membrane. The middle frame shows the OCT scan at the level of the deeper choroid. 

In the third row, the left image shows ICG fluorescence at the filling of the vascular frond. The OCT in the middle highlights the elevation of the macular lesion which is raised above the surrounding tissue. The dark surround represents the vitreous. 

In the fourth row, the image shows the ICG completely filling the membrane. The OCT image captures a distorted foveal depression (dark) adjacent to the hyperreflective membrane and concentric irregular rings created as the OCT scan cuts across sequential retinal layers pushed up by the serous fluid accumulation under the macula. 

The bottom image is a B-scan OCT image obtained along the horizontal line shown in the ICG images on the left in the second, third, and fourth rows. Underneath the B-scan image, SLO images guide the observer. The image on the right is the SLO image just before the system was switched to B-scan imaging, and the one on the left is the current SLO image. This effect is achieved by turning off the vertical scanning, which prevents it from forming a true SLO image. However, it is useful in revealing lateral movements of the eye during the acquisition of the B-scan OCT image. The straightness of the vertical bright lines indicates that the eye has not moved significantly in the horizontal direction during the 0.5 seconds necessary for generation of the B-scan OCT image. In the B-scan regimen, the three images—the large B-scan and the two small C-scans underneath—are presented simultaneously on the screen). 

Occult or poorly defined choroidal neovascular membranes have been labeled such because of the difficulty in defining their boundaries with conventional fluorescein angiography. Although the addition of ICG angiography has been helpful in identifying active choroidal disease that was often obscured by nondiscriminating fluorescein leakage, late ICG leakage has been difficult to localize to anatomic structures or landmarks. Incorporation of OCT structural imaging onto ICG angiography provides an anatomic frame of reference for late leaks and reveals the flow characteristics of specific anatomic features. 

In the following figures, the images produced by the three channels are displayed simultaneously in a four-up configuration with a fourth channel that overlays the OCT C-scan image on the ICG image. 

Figure 3shows a case (Table 1 , patient 22) of age-related macular degeneration with an occult choroidal neovascular membrane. Figure 3adisplays C-scans only, as provided in real time by the instrument. The top left frame is an SLO image showing an irregularly circular elevation in the center of the macula. The top right image is the matching coronal OCT image revealing concentric lines produced by the serous macular elevation. This hyperreflective bull’s eye formation, characteristic of serous retinal detachments in coronal OCT, is produced by the alternating hyper- and hyporeflective internal retinal layers. The bottom left frame shows the perfusion abnormalities of the macula at 33 seconds after the injection of the ICG dye. A faintly hyperfluorescent wedge-shaped area deep to the retinal vessels is evidence of a choroidal neovascular membrane. The bottom right image is a composite overlay of the top right and bottom left frames showing the spatial association between the elevation and the vascular anomaly. Overlaying of morphology and vascular structure is facilitated by the use of color, which helps distinguish the different contributions of each imaging modality. Such a superposition is allowed by the pixel-to-pixel correspondence between the simultaneously acquired SLO and ICG channels on one side and the OCT channel on the other. 

Information on the cross-sectional morphology of the retina is achieved by switching the system into the B-scan regimen. An OCT image and the image generated by the SLO channel obtained in the B-scan regimen are shown in the right column in Figure 3b . The B-scan OCT image was not calculated from C-scans but was recorded in B-scan mode, as a stack of T-scans collected along a single line for different depth values. 

In Figure 3b , the images in the left column are the C-scans in the SLO channel (top) and in the ICG channel (bottom) generated just before switching the instrument to the B-scan regimen. The bottom left frame shows the distribution of ICG dye retained in the retinal vessels and the choroidal neovascular membrane at 2 minutes 33 seconds. The top left frame shows an SLO image with circular elevation of the central macula. In the right column, the top right frame shows a longitudinal slice (B-scan) taken through the vertical red line in the SLO image. There is an irregular hyperreflective granular lesion at the level of the retinal pigment epithelium on the left side of the base of the serous cavity which corresponds to the wedge-shaped collection of dye in the ICG channel. The bottom right frame is the image provided by the SLO channel in the B-scan regimen. The only information provided by the SLO channel in the B-scan regimen relates to monitoring of movement artifacts, which are similar in both images: SLO and OCT. In this regimen, the frame scanner is stationary on a position determined by the red line in the SLO image (top left image). Therefore, the same T-scans (if there are no eye movements) are collected while the depth coordinate is scanned in the OCT channel. The depth adjustment has no relation to the SLO channel and therefore the near-perfect alignment of the lines in this image suggests minimal movement artifacts. Movement of the eye laterally toward one direction would result in horizontal deviations in the bright line in the bottom right image in Figure 3b . 

Choroidal neovascular membranes have been imaged before with ICG ²⁰ as well as with OCT (Khan J. IOVS 2004;45:ARVO E-Abstract 2979) ²¹ and we present images collected with our system, which provides such images simultaneously. An example of a case of a classic choroidal neovascular membrane in a patient with angioid streaks (Table 1 ; patient 6) is presented in the C-scan regimen in Figure 4 . The top left frames in Figures 4a and 4bare similar and show the confocal fundus image, which reveals the vascular-like pattern of cracks in Bruch’s membrane deep to the retinal vessels. There is a central, irregular, round lesion that appears to emanate from one of the streaks. In Figure 4a , the top right image is the coronal OCT, which shows a hyperreflective bull’s eye-shaped lesion indicative of a serous elevation of the sensory retina. The bright circle is the retinal pigment epithelial (RPE) layer. The bottom left frame in Figure 4ashows vascular filling by ICG and a central hyperfluorescent spot surrounded by a darker ring suggesting a vascular membrane. The bottom right image in Figure 4ais a composite overlay of the OCT and ICG demonstrating the relationship between the vascular structure of the membrane and the surrounding elevation. Also in Figure 4b , the top right frame shows a cross-sectional B-scan OCT image through the central lesion, demonstrating the vascular structure protruding through the RPE layer and the displacement of the overlying retina. The bottom right frame in Figure 4bshows the lines of the SLO image taken simultaneously with the top right frame. 

In Figure 4b , the images in the left column represent the C-scans in the SLO channel (top) and ICG channel (bottom) generated just before switching the instrument to the B-scan regimen. 

The bottom left frame in Figure 4bshows the ICG presence late in the sequence (13 minutes, 8 seconds). There is retention of the dye in the lesion but not in the rest of the retina. 

Figure 5presents sets of images from an eye with a minimally classic choroidal neovascular membrane beneath a serous retinal elevation (Table 1 , patient 14). Figures 5a and 5bshow the coronal aspect in a four-panel display: the SLO (top left), ICG fluorescence (bottom left), C-scan OCT (top right), and the overlay channel featuring a superposition of C-scan OCT and ICG images (bottom right). The images in Figures 5a and 5bwere taken at two different times in the postinjection phase. In 30 seconds, 60 of such four-up sets of images were acquired while the depth was explored over a range of typically 1.2 mm in retinal tissue. If the eye moved considerably during the acquisition and essential parts from the retina volume were missing, the acquisition could be repeated after the ICG bolus had passed. The pixel-to-pixel correspondence between the three images allowed later association of morphologic features between the two images. Generally, just a few correct en face OCT images from the stack collected during the bolus passage are sufficient for subsequent transverse alignment of any other pairs of images. The ICG fluorescence image on the left in Figure 5breveals the location of active leakage within the lesion. 

Information on the depth resolved morphology of the retina in these volumes is acquired by switching the system into the B-scan regimen. Regions of leakage, visible in the ICG image, can be selectively targeted for acquiring B scan cross sections in the OCT channel, as seen in Figures 5c(right column) and 5d. In Figure 5c , the C-scans on the left were the last images collected in the OCT and ICG channels just before switching the instrument to the B-scan regimen. The lateral size was the same as that of the C-scan images, whereas the vertical axis was oriented along the depth coordinate. 

Dedicated software was developed to position the B-scan image above the SLO or the ICG image along the line shown in the C-scan images (either SLO or ICG). Figure 5dshows the cross section OCT image intersecting the ICG image. 

Figure 6further demonstrates the utility of the system in evaluation of a retina of a patient with polypoidal choroidal vasculopathy, a peculiar variant of neovascular macular degeneration (Table 1 , patient 23). Figures 6a 6b 6cpresent the coronal aspect captured at 15, 28, and 60 seconds after the ICG injection. 

Figure 6dshows the display in the B-scan regimen, captured at 3 minutes and 18 seconds after the injection. The C-scans on the left were the last images collected in the OCT and ICG channels just before switching the instrument to the B-scan regimen. The lateral size is the same as that of C-scan images while the vertical axis is oriented along the depth coordinate. 

The ICG angiography images reveal a leash of abnormal vessels that originate near the nerve, extend inferotemporally, and terminate with bulbous endings. The accompanying OCT images capture the sausage-shaped cuff of fluid surrounding the vessels, which accounts for the lumpy polypoidal appearance in the SLO images in the top left quadrants of the display and in the cross-sectional B-scan OCT in Figure 6d(top right). The overlay channel in Figures 6a 6b 6c(bottom right) highlights the relationship between the vascular structures and the anatomic effect of the fluid leakage by displaying the vessels within the resulting cavitations. 

Discussion

In this study, we demonstrated the simultaneous operation of three separate acquisition and display channels: OCT, SLO, and ICG fluorescence in a retinal imaging system. We also showed the utility of displaying a controllable mixer channel that can mix the OCT and the ICG information (both C-scans). By software, B-scan OCT images can also be superimposed on C-scan SLO or ICG images. Clinical information revealed by each channel in part was documented for each patient in Table 1 . The pixel-to-pixel correspondence inherent in the design of this system allows an integrated and potentially more accurate analysis of the association between morphology and function within the retina and choroid than is currently possible with separate instruments. 

In these images, blood vessels are well defined in the ICG images while inconsistently revealed within the OCT images. At the same time, the depth resolution in the ICG channel is orders of magnitude lower than the OCT axial resolution and the morphology cannot be assessed accurately. Therefore, we believe that such a system can have valuable applications by combining the complementary information supplied by the two data channels. Regions of leakage, visible in the ICG image can be selectively examined in depth by acquiring B scan cross sections in the OCT channel. 

The main limitations we encountered were in imaging patients with poor optical media due to cataract, corneal scarring, or vitreous opacities. We also found that patients with poor fixation, nystagmus, or those who simply could not sit still presented particular challenges to the operator. Speed of acquisition with this system, although adequate for most patients, was a limitation for some of our more difficult cases. Control of movement artifact is, of course, the primary attraction to spectral-domain-based OCT instruments. 

The compound information acquired with the OCT and fluorescence channels is different, depending on the dye used. A combined OCT/ICG instrument is able to match the longer fluorescence wavelength to achieve better penetration into the choroid, which should enhance imaging of deeper sub-RPE disease, especially with matching combinable coronal-oriented channels. 

ICG angiography and OCT appear well suited to operate together because they share similar spectral bands. The most widely used band for OCT of the retina is 800 to 1060 nm, whereas ICG is usually excited at 806 nm and fluoresces in the 810- to 860-nm band, with a peak at 830 nm. Operating in similar bands allows the same source to be used for ICG excitation as well as for the production of an OCT image. However, the proximity of the excitation wavelength to the fluorescence band raises several technical optimization issues. For this reason, it is easier to combine OCT with fluorescein angiography, as the excitation wavelength and the fluorescence band are well separated. However, this method requires two optical sources. ¹² Such a combination is already available commercially, incorporated into the Spectralis instrument (Heidelberg Engineering). ¹⁰ A combined OCT/fluorescein angiography instrument must use a separate source to stimulate the shorter fluorescence band since the one for the OCT has to operate in the 800- to 1060-nm range for best performance in imaging the retina. Although combining these two sources may present some problems with optical alignment, it allows the matching of superficial morphology in the OCT channel to the information provided by shorter wavelength fluorescence. 

Overall, there is more clinical experience with fluorescein angiography in terms of interpretation. However, the tendency of fluorescein to spread limits its value in distinguishing occult neovascularization and identifying active leakage in areas of scarring, beneath retinal pigment epithelial detachments, and in cases of polypoidal choroidal vasculopathy. In these types of cases, combined ICG/OCT imaging would have distinct advantages. 

In summary, the value of combined OCT/ICG imaging appears best suited to the management of occult disease. The enhanced penetrating properties of ICG over fluorescein are well documented in other ICG studies. Yannuzzi et al., ²² in the days of ICG-guided laser surgery, reported that ICG permitted treatment in twice as many instances in occult choroidal neovascular members due to improved detection of vessels in the face of confounding fluorescence artifacts seen with fluorescein angiography. SLO-based ICG is particularly effective in finding small CNV lesions masked by fluorescein staining—especially relevant today, as the morbidity caused by therapy has decreased with the introduction of antiangiogenic drugs and the importance of identifying any form of active neovascularization has become the trigger for initiation of nondestructive treatment. 

³  Contributed equally to the work and therefore should be considered equivalent authors.

 Supported by the Leon Lane-sponsored ARIBA Fund of the Department of Ophthalmology, the New York Eye and Ear Infirmary (RR, GD), Ophthalmic Technologies Inc., Toronto, Canada (RR, GD, AGP), and Superlum, Moscow, Russia (AGP).

 Submitted for publication February 8, 2008; revised August 18, 2008; accepted December 19, 2008.

 Disclosure: R.B. Rosen, OTI Canada (F, C), P; M. Hathaway, OTI Canada (E, F), P; J. Rogers, OTI Canada (E, F), P; J. Pedro, OTI Canada (E, F), P; P. Garcia, OTI Canada (F), P; G.M. Dobre, OTI Canada (F), P; A.Gh. Podoleanu, OTI Canada (F), P

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

General set-up of the combined OCT/SLO/ICG- system. MX, MY: galvanometer mirrors of the x–y scanning pair. The confocal channels display ICG fluorescence and standard SLO images.

Table 1.

 

Findings as Inferred from Each Channel for the 30 Patients Imaged

Patient	Age/Sex	Diagnosis	SLO	OCT	ICG	ICG/OCT	Comments
1	22/M	Choroiditis, multifocal OS	Atrophic macular patches	Normal foveal profile	No late leaks	No hot spots to compare	ICG-OCT reveals no active leakage, early shots limited infusion error
2	66/M	PDR, CSME OU	ERM, dot hemorrhages	Cystic changes, exudates; SRF on B-scan	Multiple leak points	Hot spots within edema	OS ok, OD fuzzy; ICG-OCT reveals leaks in mound; topography not possible
3	62/M	ERM, pseudohole OD	ERM pucker	ERM, SRF; edema on B-scan	No vascular leakage	Distortion without leaks	ICG-OCT displays traction but no source of fluid, fuzzy images
4	91/F	AMD wet with hemorrhage OD	Large hemorrhage	RPED in C-scan, CNV, cysts and SRF on B-scan	CNV revealed beneath hemorrhagic mound	CNV Vascular Net within hemorrhagic mound	ICG-OCT reveals CNVM below blood, poor fixation
5	82/M	AMD wet, cicatricial	Scar, hemorrhage	Scar, no edema	CNVM vessels network	Leakage within scar	Leakage revealed as source of blood
6 (Fig. 4)	29/M	Angioid Streaks with CNV OD, inactive scar OS	Macular hemorrhage	Macular elevation	Sub-foveal CNVM	Localized leak in macular edema	ICG-OCT reveals vascular source within mound
7	74/M	AMD wet, cicatricial	Scar, lumpy surface	Subfoveal scar, with multiple irregular RPEDs	Central blockage, no leakage	RPEDs without leaks	ICG-OCT reveals inactive IPCV
8	30/M	Trauma, choroidal ruptures	Parallel choroidal ruptures	Needle-shaped RPE elevations/ruptures	Early transmission hot spots, no late leaks	Scars without active leakage	ICG-OCT confirms no activity
9	83/M	AMD	Dark macula, vitreous floaters	ERM, concentric fluid over CNV	Late central leakage	SRF over RPED and CNVM	ICG-OCT reveals chronic CNVM beneath SRF/RPED mound
10 (Fig. 2)	76/M	AMD first patient on system	Three channel display	Concentric fluid over CNVM	Feeder vessel frond	Mixing channel not implemented	Two-channel study, feeder vessel frond source of n macular edema
11	76/M	AMD wet macular hemorrhage OD after treatment	Fibrotic mound	Scar, subretinal fluid, cystic changes	CNVM vessels network	CNVM seen within scar	ICG-OCT reveals mature CNVM within fibrosis with SRF
12	76/M	AMD, later post-treatment study	Fibrotic mound	Scar, no SRF, RPED	CNVM vessels network	CNVM in scar, no leakage	ICG-OCT shows no residual leakage
13	76/F	AMD wet, scar OD r/o active OS	Cicatricial mound	CME over CNVM with fibrosis	CNVM vessels network	CME over CNVM	ICG-OCT reveals chronic CNVM beneath CME in mound
14 (Fig. 5)	82/M	AMD wet	Pigment changes, lumpy	SRF, CNVM, multiple RPEDs	Focal/plaque leaks	RPEDs, leaks in mound	ICG-OCT shows RPEDs and leakage sources within macular mound
15	73/F	AMD	Macula edema wrinkling,	Outline of CNVM, RPEDs	Feeder vessel, CNVM	CNVM within elevation	ICG-OCT reveals feeder vessel, CNVM within mound
16	59/F	IPCV	Peripapillary mound OS	RPED with trace SRF next to disc,	Faint staining of mound	Late staining of polypoidal mound	ICG-OCT reveals inactive IPCV
17	24/M	Myopic degn. CNV, PIC	Focal edema, hemorrhage	Eccentric macular elevation	Eccentric leakage	Localized leak in macular edema	ICG-OCT reveals localized CNVM in scar eccentric to fovea
18	41/M	Atypical CSR OS vs CNVM	Eccentric focal edema OS	Small SRF with extrafoveal RPED	No leaks	Elevation but no vessel leaks	ICG-OCT confirms non-vascular source of elevation, atypical CSR
19	82/F	AMD Wet OS	Focal edema, hemorrhage	Polypoidal RPED	Irregular faint leak	Leak in polypoidal mound	Lesion poorly defined due to cataract
20	48/M	PDR OU; VH OS	Dot heme, no edema OD	Good foveal profile, no CSME	Focal microaneurysm hot spots	Hot spots without edema	ICG-OCT shows microaneurysm leaks without CSME
21	77/M	Macular serpiginous OS	Pigment changes	RPE disturbance, no elevation	Wedge of telangiectatic vessels	Minimum leaks within atrophic area	ICG-OCT reveals leaks without edema, cicatricial changes
22 (Fig. 3)	68/F	AMD r/o CNVM	Serous macular elevation	Concentric fluid rings	CNVM at RPED edge	CNVM at edge of SRF	ICG-OCT reveals occult CNVM beneath SRF over RPED, little fuzzy;
23 (Fig. 6)	68/F	IPCV OS	Multiple RPEDs	Outline of RPEDs	Focal leaks	Leaks localized in RPEDs	ICG-OCT reveals focal leaks within multiple polypoidal lesions
24	63/M	Chronic RPED OU	Pigmentary Changes, lumpy	RPEDs in clusters, noSRF	Hypofluorescent, no leaks	Nonvascular RPEDs	ICG-OCT reveals no leakage source within multiple RPED lesions
25	70/F	AMD OD after macugen	Blood, edema next to scar	Cystic mound, SRF next to scar	CNVM at edge of scar	CNVM within elevation	ICG-OCT reveals new CNVM in edema at edge of scar
26	83/M	AMD with CNV OS	Pigment changes, lumpy, macula	Multiple RPEDs concentric SRF	Multiple leaks	Leaks localized in RPEDs	ICG-OCT reveals CNV associated with RPED/SRF mounds
27	49/M	CSR OU	Pigment change, elevation OU	SRF, RPEDs OU	Diffuse staining OU	Chronic leakage, no scars	ICG-OCT reveals chronic leaks of CSR, no evidence of CNVM
28	36/F	CSR OD vs vitelliform lesion	Foveal disturbance	Subfoveal serous detachment	No focal leaks seen	No hot spots associated with SRD	ICG-OCT reveals no focal leakage related to macular elevation
29	85/M	RPED, occult CNVM OD	RPED OD	RPED with Irregular border	Peripheral leak	CNVM in notch	ICG-OCT shows CNVM at edge of RPED, study limited by cataract OD
30	73/F	AMD Wet OD	Crescent-shaped lesion	Subfoveal mound	CNVM vascular frond	CNVM within elevation	ICG-OCT reveal CNVM within fibrous mound

 CNVM, choroidal neovascular membrane; Degn, degeneration; ERM, epiretinal membrane; RPED, retinal pigment epithelial detachment; PIC, punctate inner choroidopathy; SRD, serous retinal detachment; SRF, subretinal fluid.

Figure 2.

 

Choroidal neovascular membrane in a poorly defined macular lesion. Top: color fundus image; Next three rows: ICG fluorescence (left), en face OCT (middle), and SLO (right) images of the eye fundus; bottom: B-scan OCT image obtained along the horizontal line shown in the ICG images. C-scan images: 12° × 12°. B-scan: lateral size: 12°, depth 1.3 mm measured in air.

Figure 3.

 

Age-related macular degeneration with occult choroidal neovascular membrane. (a) Coronal aspect, the four channel images as displayed in real time. Top left: SLO; bottom left: ICG; top right: OCT; bottom right: overlay of OCT and ICG images. Size: 12° × 12°. (b) Compound display of C-scans and B-scans. Left column, top: SLO; bottom ICG image. Size: 12° × 12°; top right: B-scan OCT image, lateral size 12°, depth 1.3 mm in air. Bottom right: the SLO image deprived from vertical scanning acquired during the B-scan OCT imaging. The relative verticality of bright line documents the minimal lateral movement of the eye during the 0.5-second acquisition. The squaring of the aspect ratio of the B-scan image results in a slight vertical exaggeration of the retinal elevation.

Figure 4.

 

Classic choroidal neovascular membrane in a patient with angioid streaks. (a) Coronal aspect, the four channel images as displayed in real time. Top left: SLO; bottom left: ICG; top right: OCT; bottom right: overlay of OCT and ICG images; size, 12° × 12°. (b) Compound display of C- and B-scans. Left column, top: SLO; bottom: ICG image; size, 12° × 12°. Top right: B-scan OCT image: lateral size, 12°; depth, 1.3 mm in air. Bottom right: the SLO image deprived from vertical scanning acquired during the B-scan OCT imaging. The perfect verticality of bright lines indicate that the eye did not move laterally during the 0.5-second acquisition.

Figure 5.

 

SLO/OCT/ICG image sets acquired from a patient with minimally classic choroidal neovascularization. (a) SLO (top left), ICG fluorescence (bottom left), en face OCT (top right), and OCT/ICG overlay (bottom right) images in the preinjection phase at 1.5 seconds. (b) Same set of channels shown after injection at 8.5 seconds. (c) Four-image set displayed in the B-scan regimen. SLO (top left), ICG fluorescence (bottom left), B-scan OCT (top right), and confocal lines (bottom right). The red line on the SLO image is the line where the B-scan OCT crossed the coronal slice. The confocal lines in the bottom right show any movement artifacts during the acquisition of the OCT image. The C-scan images on the left were acquired before switching the instrument from the C-scan to the B-scan regimen. (d) 3-D rendering of the intersection of the ICG (C-scan) and the OCT plane (B-scan) showing the correspondence between areas of leakage and overlying RPE elevation. In all C-scan images: lateral size, 12° × 12°; B-scan OCT images: lateral size, 12° and 1.3 mm in depth measured in air.

Figure 6.

 

ICG/OCT/SLO sets of a patient with polypoidal choroidal neovascularization. (a) Early arterial phase of ICG sequence reveals lobular choroidal vessels. OCT depth is within the choroid and shows evidence of shadowing. (b) Mid arterial-venous phase demonstrates leash of deep abnormal vessels with hyperfluorescent bulbous endings. OCT image outlines the overlying serous elevation surrounding the vessels and hot spots. (c) Full venous phase of the ICG angiogram shows increased leakage at vessel endings. The OCT reveals the outlines of the serous cuff around the vessels and enlarging fluorescence accumulations. (d) Late-phase ICG with B-scan OCT through the area of leakage. The OCT reveals corrugated elevation of the RPE. The SLO image in the bottom right confirms good alignment, with minimal movement artifacts. The z-axis of the B-scan is expanded by the configuration of the multichannel display producing an exaggeration of the aspect ratio. In all C-scan images, lateral size, 12° × 12°; B-scan OCT image, lateral size, 12° and 1.3 mm in depth measured in air.