i m a g i n g

r e v i e w

Enhanced Depth Imaging Optical

Coherence Tomography
Ian Y. Wong, MRCS, FHKAM (Oph); Hideki Koizumi, MD, PhD;
Wico W. Lai, MD, FHKAM (Oph)

ABSTRACT INTRODUCTION

Imaging the choroid with conventional commercial The introduction of optical coherence tomogra-
spectral-domain optical coherence tomography (SD- phy (OCT) has revolutionized the understanding of
OCT) has been difficult, mainly because of difficulty in the eye.1-4 It has the advantage of providing the user
signal transmission beyond the retinal pigment epithe- an in vivo cross-sectional image of the retina, which
lium. A recent modification to the standard technique, could not be achieved with other means of imaging.
termed enhanced depth imaging optical coherence to- A detailed outline from the inner vitreoretinal inter-
mography (EDI-OCT), was able to image the choroid face to the outer retinal pigment epithelium could be
with reasonable clarity using commercial SD-OCTs. visualized. Presence of vitreoretinal adhesion, retinal
The aim of this article was to review the technique, thickening, and intraretinal or subretinal fluid collec-
principle, recent findings, and possible future develop- tion can be seen with high precision. Quantification of
ments regarding EDI-OCT. A MEDLINE search on retinal thicknesses is also possible where serial changes
all published articles on EDI-OCT was performed up may be useful in documenting progression of disease,
to December 2010. The principle behind EDI-OCT either in retinopathies or glaucoma.5-7
was discussed. Modification to the conventional tech- The choroid, which accounts for most ocular
nique in image acquisition was described and illustrated blood flow,8 may be affected in several disease states,
with figures. EDI-OCT findings in various retinal and such as polypoidal choroidal vasculopathy and cho-
choroidal diseases were discussed. Advantages and dis- roidal melanoma. It is also prone to suffer from age-
advantages were also discussed. EDI-OCT has proved related degeneration, microvascular artherosclerotic
to be a promising novel technique in imaging the cho- changes, and changes inherent to other microvascular
roid. [Ophthalmic Surg Lasers Imaging 2011;42: systems.9-12 An understanding of the choroid using
S75-S84.] non-invasive imaging techniques has been limited. For

From the Eye Institute (IYW, WWL), Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong; and the Department of Ophthalmology (HK),
Kyoto Prefectural University of Medicine, Kyoto, Japan.

Originally submitted January 17, 2011. Accepted for publication February 25, 2011.
The authors have no financial or proprietary interest in the materials presented herein.
Address correspondence to Ian Y. Wong, MRCS, FHKAM (Oph), Room 301, Level 3, Block B, Cyberport 4, 100 Cyberport Road, Hong Kong. E-mail:
ianyhwong@gmail.com
doi: 10.3928/15428877-20110627-07

Ophthalmic Surgery, Lasers & Imaging Vol. 42, No. 4 (Suppl), 2011 S75
 i m a g i n g

Figure 1. Normal appearance of the retina

and choroid at the same location as imaged
with (A) conventional protocol versus (B)
enhanced depth imaging optical coherence
tomography (EDI-OCT). Images were taken
using the Heidelberg Spectralis (Heidelberg
Engineering, Heidelberg, Germany). More
detail of the choroid could be visualized
with EDI-OCT compared to the conventional
method. At the same time, retinal details
could be seen with comparable quality.

instance, imaging of the choroid with OCT was not arm. The interferogram obtained will then undergo Fou-
entirely possible.13 This was because the wavelength of rier transformation to form scattering amplitudes. This
the light source used to image the retina was not long will then be analyzed and reformed into the retinal image
enough to penetrate into the choroid.14 This was due obtained on the capturing screen. In SD-OCT, the real
to wavelength dependent light scattering and signal image is always accompanied by an inverted image.17-19
loss that occur in the image path, decreased sensitivity However, only one of the two images is generally shown
away from zero-delay, and the various inherited limi- on the capturing screen (ie, the upright real image). In the
tations with Fourier transformation.15 The common real image, the inner part of the retina is shown facing up
commercially available OCTs employ a wavelength of and the choroid is shown at the bottom of the screen.
800 nm, whereas those capable of choroidal imaging The reasons why the choroid cannot be imaged clear-
are in the range of 1,060 nm.14,16 One would have to ly could be attributed to (1) decreasing sensitivity and
use a light source nearer to the infrared region to pen- resolution with increasing displacement from zero-delay,
etrate into the choroid. This was not possible for most (2) decreased maximal dynamic range inherent in Fourier
commercially available OCTs, because clarity for reti- domain systems, (3) wavelength dependent light scatter-
nal structures would be compromised. Following the ing and signal loss in the image path, and (4) the lateral
introduction of a new imaging technique, referred to width of the defocused imaging beam.13 When the instru-
as enhanced depth imaging optical coherence tomog- ment is moved closer to the eye to image deeper layers, the
raphy (EDI-OCT), choroidal imaging with standard inverted image is displayed such that the choroid is shown
commercially available spectral-domain OCT (SD- facing up (ie, closer to zero-delay) while the inner retina
OCT) was made possible.13 This review provides the is facing down. This has the effect of delivering the most
reader with an up-to-date account of this technique closely focused portion of the illumination at the level of
and the various findings published to date. the choroid or the inner scleral border. Hence, the choroid
can be imaged at higher sensitivity (Fig. 1).
METHOD The wavelength employed by most SD-OCTs is able
to give a relatively low signal-to-noise ratio and, by aver-
A literature search on MEDLINE was performed aging more frames, a relatively clear and noise-free im-
on key words optical coherence tomography, en- age could be obtained. Therefore, EDI-OCT is a simple
hanced depth imaging, spectral domain OCT, and modification of the conventional SD-OCT technique,
choroid up to December 2010. Search results were made possible by a slight displacement of the image cap-
limited to those published in English and studies on turing machine, and computerized image averaging.
human subjects only.
Standard Procedure
Principle of EDI-OCT There are several SD-OCTs available commercially,
In SD-OCT, a broadband light source is employed but to date, only the Heidelberg Spectralis OCT (Heidel-
to image the retina. Interference signal is then generated berg Engineering, Heidelberg, Germany) and the Cirrus
by comparing the signal received to that from a reference HD-OCT (Carl Zeiss Meditec Inc., Dublin, CA) have

S76 Copyright SLACK Incorporated

 i m a g i n g

Figure 2. Enhanced depth imaging scans

through the same location across the fo-
vea: (A) image obtained after averaging 20
frames and (B) image obtained after averag-
ing 100 frames. Note the lower noise-signal
with more frames averaged, as shown in (B).
Differentiation of the outer retinal pigment
epithelium/choroid border, and the choroid
scleral junction is still possible even by aver-
aging lesser frames.

Figure 3. Appearance of a 6-mm horizontal

raster line scan through the fovea in a normal
subject. Arrow = junction between the retinal
pigment epithelium and choroid; triangle =
junction between the choroid and sclera.

Figure 4. Manual measurements of choroi-

dal thickness at each 1,000-m interval, cen-
tered on the fovea. Calipers were positioned
manually using computer software provided
by the manufacturer.

been described to be capable of performing EDI-OCT.13,20 porates EDI-OCT as an added feature. The software
Here we explain the principle using the Heidelberg Spec- update (version 5.3) automatically inverts the captur-
tralis OCT, as previously described by Spaide et al.13 ing screen and positions zero-delay inferiorly instead
After positioning the patient and achieving focus, of superiorly. In doing so, the operator can acquire the
the OCT machine is slightly pushed toward the patient EDI-OCT images with increased ease because now
such that an inverted image is obtained on the capturing the image is upright, rather than inverted. However,
screen. Because image quality is best toward zero-delay this feature is only available in the updated Heidelberg
(ie, top of the screen), the inverted image is captured Spectralis software, whereas for the Cirrus HD OCT,
when it is as close to the top as possible. Special attention acquisition using the inverted-image technique is still
is needed to avoid capturing a folded image (ie, when required. Despite advancement in image acquisition,
the machine is not close enough to the eye and the image automated choroidal thickness measurement is still un-
has not yet been inverted completely). To enhance clar- available at the time of writing this review.
ity and reduce noise signal, each image should be aver- Image processing is done with the standard Heidel-
aged from 100 frames. Eye-tracking technology also im- berg Explorer software. Because no automatic measuring
proves image quality and acquisition time (Fig. 2). The software for EDI-OCT exists, manual measurement has
appearance of an EDI-OCT image of a normal line scan to be done on each image (Fig. 4). Choroidal thickness
through the fovea is shown in Figure 3. is defined as the distance between the outer border of the
A recent software update by Heidelberg incor- retinal pigment epithelium and the inner scleral border

Ophthalmic Surgery, Lasers & Imaging Vol. 42, No. 4 (Suppl), 2011 S77
 i m a g i n g

Table 1 and metabolic activity. The metabolic demand has to

Choroidal Thickness (m) as Measured be met by the choroidal circulation because there is
With Different Spectral-Domain Optical no direct blood supply from the retinal circulation in
Coherence Tomography the fovea. It is therefore reasonable to find the thickest
part of the choroid to be located underneath the fovea,
Mean Choroidal Heidelberg Cirrus HD
Thicknessa Spectralis21 OCT23 where the demand for oxygen is highest.
Temporal 3.0b 261 n/a
Correlation With Age
Temporal 2.5 266 218
Margolis and Spaide also correlated choroidal
Temporal 2.0 268 237 thickness with age.21 It was found that the choroid gets
Temporal 1.5 268 248 thinner with age and the correlation was statistically
Temporal 1.0 273 256 significant at all points measured on the 6-mm line.
Temporal 0.5 277 264 With regression analysis, for instance, subfoveal cho-
Fovea 287 272 roidal thickness was found to decrease by 1.56 m for
Nasal 0.5 276 261
each year of age. The choroid, being a vascular struc-
ture, is prone to be affected by systemic conditions such
Nasal 1.0 253 248
as hypertension and hyperlipidemia, and it is likely to
Nasal 1.5 232 217 undergo artherosclerotic and aging changes similar to
Nasal 2.0 203 186 other microvascular structures within our body.8 In a
Nasal 2.5 170 157 previous study on eye bank and autopsy eyes, the cho-
Nasal 3.0 145 n/a roidal thickness was found to correlate negatively with
n/a = data not available. age (decrease 1.1 m per year of age).22 Results from
a
Mean choroidal thickness as measured from outer border of histological examination may not necessarily agree with
retinal pigment epithelium to inner scleral border.
b
Temporal 3.0 refers to measurement point 3.0 mm temporal to in vivo findings, mainly because (1) the tissue fixation
fovea on a horizontal line scan.
Heidelberg Spectralis is manufactured by Heidelberg Engineer-
method may have altered tissue texture and hence re-
ing, Heidelberg, Germany. Cirrus HD-OCT is manufactured by sults may be altered and (2) choroidal thickness also
Carl Zeiss Meditec, Inc., Dublin, CA.
reflects circulation but circulation ceases in diseased tis-
sue, which may interfere with interpretation.13,21 De-
(Fig. 3). Manual adjustments have to be done on the spite that, in vivo findings obtained by Margolis and
image captured. However, reproducibility was relatively Spaide confirmed that choroidal thickness measured in
good and measurement was possible in most cases.13,20 vivo follows the same pattern.21

FINDINGS IN NORMAL SUBJECTS Measurements With Other SD-OCT

Margolis and Spaide did their measurements with
Thickness Profile Along the Macula the Heidelberg Spectralis. Manjunath et al. adopted
In a pilot study by Margolis and Spaide, 54 nor- their protocol and produced results with another mod-
mal eyes were studied.21 The choroidal thickness was el of SD-OCT, the Cirrus HD OCT.23 The differences
measured at 500-m intervals along a 6-mm horizon- between the Spectralis and Cirrus HD OCT were that
tal line centered on the fovea (ie, measurements were Cirrus lacks eye-tracking ability and it can only per-
made from 3 mm nasal to the fovea to 3 mm temporal form 20 B-scans at a time for each measurement. In
to the fovea).21 Results showed that the choroid was contrast, Spectralis is superior in that it has eye-track-
thickest underneath the fovea, with a mean of 287 m. ing ability and can obtain 100 B-scans simultaneously.
Choroidal thickness decreased on both sides of the fo- Despite that, Manjunath et al. were still able to produce
vea, but more so nasally, and was found to be thinnest comparable results in terms of variation in choroidal
at the nasal end of the imaging line (ie, 3 mm nasal to thickness along the macula and the negative correlation
the fovea). The choroid is the most vascular structure with age.23 Table 1 outlines the measurements of cho-
within the eye,8 and the fovea, situated at the center roidal thicknesses obtained using these two models of
of the macula, has the highest photoreceptor density SD-OCT. To date, the Spectralis and Cirrus HD OCT

S78 Copyright SLACK Incorporated

 i m a g i n g

are the only commercial models of SD-OCT that were

reported to be capable of performing EDI-OCT.

FINDINGS IN PATHOLOGICAL CONDITIONS

Myopia
Fujiwara et al. studied the EDI-OCT pattern of
choroidal thickness in subjects with myopia of 6 diop-
ters or more.24 A 6-mm horizontal scan line was used,
centered on the fovea. Choroidal thickness was mea-
sured at 1,000-m intervals along the line. The choroid
was found to be thinnest at the nasal end. Thickness
increased in a graded fashion toward the temporal side
and reached maximum at the temporal end. This was in
contrast to that in normal subjects, where thickness was
highest under the fovea.21,23 A possible explanation was
the presence of posterior staphyloma, which may have
distorted the normal architecture of the choroid.24 De-
spite that, choroidal thickness correlated negatively with
age and regression analysis showed a 1.27-m decrease
with each year of age. This was similar to that found in
normal subjects.21,23 Furthermore, choroidal thickness Figure 5. Segmental imaging of pigment epithelium detachment
was found to decrease with increasing myopia. An 8.7- (PED) with Heidelberg Spectralis (Heidelberg Engineering, Hei-
m decrease was found for each diopter of myopia.24 delberg, Germany) using enhanced depth imaging (EDI-OCT).
Fujiwara et al. further subdivided the subjects into Heterogeneity within the PED could be seen when EDI-OCT was
those who had myopic choroidal neovascularization employed. (A) Image of PED taken with conventional method.
Relative hollowness of the PED is seen. (B) Same section through
(CNV) and those who did not.24 Results showed that
PED images with EDI-OCT. Contents within PED better shown.
in those who had had CNV, the choroid was thinner Band of hyperreflective tissue seen overlying Bruchs membrane,
when compared to those who did not and the contour as outlined by arrowheads, was not seen clearly in (A).
was distorted. This was because of either the effect of
the CNV or the effect of treatment (ie, photodynamic PEDs.26 It was thought that the Bruchs membrane turns
therapy or anti-vascular endothelial growth factor in- increasingly hydrophobic and impermeable to fluid with
jection), or a combination of both.24 age, and fluid would accumulate underneath the retinal
pigment epithelium during the normal outflow pathway.
Age-Related Macular Degeneration Hence a PED would form if enough fluid accumulates
Before the description of the technique of EDI-OCT, beneath the retinal pigment epithelium, and CNV forms
imaging pigment epithelial detachment (PED) with OCT as a secondary product.27 Other theories exist, but none
was difficult. It often appears as optically empty. There- have been conclusive.28-31
fore, the pathogenesis of PEDs associated with age-related With EDI-OCT, sensitivity and resolution is en-
macular degeneration (AMD) has been controversial. hanced because the choroidal side of the image is placed
Gass postulated that choroidal neovascularization (CNV) closer to zero-delay. In a retrospective study of 22 eyes
grows under the retinal pigment epithelium and spreads with PEDs associated with AMD, Spaide was able to
laterally and causes exudation. Accumulation of exudates image the inner contents with this new technique.15
increases the pressure between the choroid and retinal In his study, the full extent of the choroid was imaged
pigment epithelium and causes ballooning and PED for- with this new technique in all eyes. The PED was filled
mation.25 On the other hand, Bird and Marshall thought with hyperreflective tissue in half of them, whereas it
that abnormal interaction between the retinal pigment was filled with serous fluid and reflective substance
epithelium and Bruchs membrane causes the formation of in the remaining half (Fig. 5). He went on to postu-

Ophthalmic Surgery, Lasers & Imaging Vol. 42, No. 4 (Suppl), 2011 S79
 i m a g i n g

late that the reflective substance represents CNV and 19 patients with CSC, they found a mean subfoveal
should be part of the PED complex, but it was not be- choroidal thickness of 505 m (standard deviation =
ing visualized before. In one particular case among his 124 m; range = 439 to 573 m). This was higher than
series, Spaide noted a reflective band underneath the normative data reported previously.21,23 Among those
retinal pigment epithelium inside the PED that also who had unilateral CSC, choroidal thickness was also
contained serous fluid. One week after treatment with increased in the disease-free fellow eye.44 Increased
ranibizumab, the band straightened and separated from choroidal thickness may represent increased circulation
the undersurface of the PED, followed by partial col- and vascular dilatation. This is in agreement with pre-
lapse of the whole PED. A month later, serous content vious indocyanine green angiography studies, in that
was absent and the PED flattened out and was totally both eyes have increased choroidal circulation even if
filled with the reflective substance, presumably CNV.15 only one eye has clinically demonstrable CSC.39 There
Spaide concluded that PED in association with AMD are several known risk factors for CSC, such as stress,
is likely neovascular in origin, and that the Gass hy- pregnancy, sympathomimetic agent use, corticosteroid
pothesis is more likely to be the case.15 use, hypertension, and autonomic dysfunction.44-49
A point worth noting is that signals in the highest These are systemic risk factors rather than local ones,
part of the PED tend to be blurred when performing and the choroid is prone to be affected because it is a
EDI-OCT in subjects with large PEDs. This is because vascular structure with no autoregulation. The results
the higher the PED is, the further away it is from zero- demonstrated by Imamura et al. further confirmed this
delay. As mentioned earlier, signals are clearest nearer finding.44
zero-delay. Because this is intrinsic to the principle of In another study on CSC with EDI-OCT, Maruko
EDI-OCT, successful capturing of large PEDs may et al. have shown that choroidal thickness is reduced af-
only be possible when future enhancement in OCT ter successful treatment with photodynamic therapy.50
technique is available, possibly by enhancing the ef- In their study, 20 patients with CSC were treated with
fective imaging depth and improving noise reduction either laser photocoagulation or photodynamic therapy.
algorithms. Pre-treatment choroidal thickness was 345 m in the
Although the new findings may lead us to a deeper laser photocoagulation group and 389 m in the pho-
understanding of the pathogenesis of PEDs, there is still todynamic therapy group. Choroidal thicknesses at 4
more to be done, histological correlation in particular. weeks were 340 m in the laser photocoagulation group
Also, further studies have to be done for other variants (P = .2) and 330 m in the photodynamic therapy group
seen in the AMD spectrum (ie, geographical atrophy or (P < .001). There was a transient increase in choroidal
classic CNV). Nevertheless, EDI-OCT proved to be a thickness during the initial 2 days in the photodynamic
promising future development to allow us to improve therapy group that decreased at 4 weeks to a lower level
the management and understanding of AMD. than that at pre-treatment. Photodynamic therapy is ca-
pable of producing choriocapillaris damage and vascular
Central Serous Chorioretinopathy remodeling in the choroid.51-59 Applying photodynamic
Central serous chorioretinopathy (CSC) is char- therapy in CSC was thought to reduce vascular hyper-
acterized by serous detachment of the neurosensory permeability and in return cessation of the pathological
retina, sometimes accompanied by a serous PED. It process.50 In contrast, those treated with laser photoco-
was thought that vascular dilatation, congestion, and agulation did not have reduction in choroidal thickness,
hyperpermeability of the choroidal vessels create an in- meaning hyperpermeability was not reduced. This was
creased hydrostatic pressure within the choroid, lead- because the area of hyperpermeability was larger than
ing to PED formation. This collection of fluid subse- the spot of focal laser applied.60,61 Although local effect
quently leaks into the subretinal space through focal was achieved around the laser spot, diffuse reduction
leakage points, causing CSC formation.32-44 The above in hyperpermeability was not observed; thus, choroidal
theories were backed up by studies using indocyanine thickness reduction was not evident. Although the study
green angiography. by Maruko et al. cannot be a guide for better treatment
With EDI-OCT, Imamura et al. have found new for CSC, it did highlight the value of EDI-OCT in fu-
supporting evidence for these theories.44 In a study of ture studies of CSC.

S80 Copyright SLACK Incorporated

 i m a g i n g

Inherited Retinal Diseases tively lower resolution in choroidal images than in con-
EDI-OCT imaging in inherited retinal diseases is ventional retinal scans and the lack of eye-tracking in
possible and correlates well with the clinical appear- some models of OCT (only the Heidelberg Spectralis is
ance of the disease. Yeoh et al. reported EDI-OCT currently able to perform eye-tracking for EDI-OCT).
imaging of the choroid in 20 patients with inherited To enhance repeatability, automated measurement
retinal diseases, including Stargardts disease, macular software should be developed for choroidal measure-
dystrophy, cone/rod dystrophy, Bests disease, Biettis ments and eye-tracking function should be ideally in-
disease, choroideremia, and bifocal chorioretinal at- corporated into every OCT machine.
rophy.62 Although the subjects were heterogeneous
in terms of etiology, some general principles were ob- DISCUSSION
served. It was found that choroidal thickness varied ac-
cording to the type of disease, and the severity of the Before the introduction of EDI-OCT, the means
underlying disease. In general, choroidal thinning was of assessing the choroid included contact ultrasonogra-
observed where visible retinal or chorioretinal changes phy,63 magnetic resonance imaging,64 and histological
were present. Choroidal thinning may be focal or dif- studies.22,65 Contact ultrasonography can provide an in
fuse depending on the underlying retinal dystrophy, vivo image of the choroid and can still function in the
and was symmetrical bilaterally in all cases, which was presence of dense media opacity (eg, cataract), but it
in agreement with the bilateral nature of the respec- has low resolution and precise location of measurement
tive inherited retinal disease. Yeoh et al. divided those is difficult.65 Magnetic resonance imaging is non-inva-
with mild to moderate thinning from those with severe sive, yet differentiation between retina, choroid, and
thinning.62 Mild to moderate thinning was thought to sclera is difficult.64 Histological studies have disadvan-
represent choriocapillaris atrophy secondary to retinal tages in that (1) the eye has to be removed for study, (2)
pigment epithelium cell death, whereas severe thinning tissue fixation may affect tissue texture, and (3) cessa-
was thought to be genetic in nature and caused atro- tion of circulation may influence the histology.22
phic changes in larger choroidal vessels. Therefore, EDI-OCT has several advantages: (1) it
Due to the retrospective nature, limited number uses a commercially available machine, (2) it is non-in-
of cases, and heterogeneity of etiologies, it was difficult vasive, (3) it provides an in vivo account of the choroid,
to allocate a specific type of choroidal thinning pattern and (4) its operation is simple and has high repeatabil-
to one particular retinal disease. However, the study ity. However, it is still limited by (1) the lack of auto-
formed a framework for future studies in this regard. mated choroidal thickness measurement software that
requires manual measurement to be done, which can
Repeatability be time-consuming and creates bias; (2) the presence
Repeatability is critical when an imaging tech- of media opacities, which hinders signal transmission
nique is to be implemented in practice. For EDI-OCT, and reception between the OCT and the choroid; and
there is no automated software to measure choroidal (3) the relative high cost of the SD-OCT machine. A
thickness. All measurements must done manually. This summary of recent important findings in EDI-OCT is
raises concern regarding inter-observer and inter-ses- given in Table 2 for reference.
sion repeatability. A dedicated automated software to measure choroi-
Spaide et al. reported good inter-observer repeat- dal thickness is clearly required. Further correlation in
ability with the Heidelberg Spectralis (r = 0.93 right normal subjects is needed, for example in the pediatric
eye, r = 0.97 left eye; P < .001).13 Rahman et al. pro- population. Correlation in pathological conditions is
duced similar results and found that a change of more also needed, especially in those of choroidal origin (eg,
than 32 m is likely to exceed inter-observer variability choroidal nevus or polypoidal choroidal vasculopathy).
(inter-observer coefficient of repeatability was 32 m,
95% confidence interval was 30 to 34 m).20 CONCLUSION
Although good inter-observer repeatability was
produced with manual measurements, there are still The introduction of EDI-OCT has provided new
difficulties sometimes. The main problems are the rela- means of assessing the choroid with commercially avail-

Ophthalmic Surgery, Lasers & Imaging Vol. 42, No. 4 (Suppl), 2011 S81
 Table 2

S82
Summary of Findings in Enhanced Depth Imaging Optical Coherence Tomography
Investigator No. Pathology Scanning Protocol OCT Machine Choroidal Thickness Other Important Findings/Remarks

Spaide et al., 17 Normal 7 sections, each comprised 100 averaged scans, Heidelberg Mean 318 to 335 m under fovea. High reproducibility (r = 0.93 to 0.97; P < .01).
200813 volunteers within a 5 3 15 degree rectangle centered on the Spectralis
fovea. Eye-tracking mode on.
Margolis & Spaide, 30 Normal 7 sections, each comprised 100 averaged scans, Heidelberg Mean 287 m under fovea. Thinnest Negative correlation with age, decrease by
200921 volunteers within a 5 3 30 degree rectangle centered on Spectralis nasal end, thickest at fovea, gradual 15.6 m each decade of life.
the fovea, at 500-m intervals from 3 mm nasal decrease in thickness from fovea to
to fovea to 3 mm temporal to fovea. Eye-track- temporal end.
ing mode on.
Spaide, 200915 22 AMD with 7 sections, each comprised 100 averaged scans, Heidelberg Not studied. In 50% of cases, PEDs were entirely filled with
PED within a 5 3 15 degree or larger rectangle to Spectralis hyperreflective tissue. In the remaining 50%
encompass the PED and associated neovascu- of cases, PEDs contained hyperreflective tis-
larization. Eye-tracking mode on. sue and serous fluid collection. Administrat-
ing ranibizumab caused PED to flatten and
hyperreflective tissue inside to contract.
i

Fujiwara et al., 31 High myopia 7 sections, each comprised 100 averaged scans, Heidelberg Mean 93.2 m under fovea. Mean Negative correlation with age, (P = .006),
200924 >6D within a 5 3 30 degree rectangle centered on Spectralis refractive error -11.9 D. decrease by 12.7 m each decade of life.
the fovea, at 1,000-m intervals from 3 mm Negative correlation with refractive error (P
nasal to fovea to 3 mm temporal to fovea. Eye- < .001), decrease by 8.7 m each diopter of
m

tracking mode on. myopia. Negative correlation with history of

choroidal neovascularization (P = .013).
Imamura et al., 19 CSC Same as Margolis & Spaide21 Heidelberg Mean 505 m under the fovea (ie, Choroidal thickness in the fellow eye had a
a

200944 Spectralis mean 214 m thicker than previously mean of 451 m, which was also higher than
reported, P < .001). that reported in normals.
Maruko et al., 20 CSC Same as Spaide et al.13 In addition, scanning Heidelberg In the LP group, mean thickness was Choroidal thickness was higher than normal
g

201050 was done before and after treatment, either by Spectralis 345 m before and 340 m after in CSC. Subretinal fluid resolved in all
LP or PDT. treatment (Not significant, P = .2). In cases after treatment. Changes in choroidal
the PDT group, mean thickness was thicknesses were noted after PDT, but not
i

389 m at baseline, 462 m at day 2, after LP. Suggested therapeutic mechanism

360 m at 1 week, and 330 m at 4 difference between LP and PDT.
weeks (P < .05 at all points).
n

Yeoh et al., 201066 20 Inherited Scanning done on areas of focal retinal thinning, Heidelberg 50% had no choroidal thinning and Choroidal thickness correlated well with clinical
retinal and adjacent areas of normal-looking retina. Spectralis mean subfoveal choroidal thickness appearance of retinopathy, and was found to
diseases Eye-tracking mode on. was 312 to 317 m. Among the other be symmetrical. Extent of choroidal thinning
50%, some degree of thinning was dependent on the stage of disease. Cho-
g

found, ranging from mild, moder- roidal thinning did not necessarily correlate
ate, to severe. Thinnest choroid was with vision.
70 m in the affected area; despite
that, adjacent area was found to be
normal.
Rahman et al., 50 Normal 2 horizontal and 2 vertical line scans through the Heidelberg Mean 332 m under fovea. Good repeatability was found. Intra-observer
201020 volunteers fovea, each averaged from 100 frames. Eye- Spectralis CR was 23, inter-observer CR was 32 to 34.
tracking mode on. A change of more than 32 m was likely to
exceed inter-observer variability.
Manjunath et al., 34 Normal 1 horizontal line scan through the fovea, Cirrus HD-OCT Mean 272 m under fovea. Thinnest Good inter-observer correlation (r = 0.92 to
201023 volunteers measured at 500-m intervals, along the line nasal end, thickest at fovea, gradual 0.93, P < .0001). Poor retinal-choroidal thick-
2,500 m temporal to 2,500 m nasal of the decrease in thickness from fovea to ness correlation (r = -0.23, P = .18).
fovea. Each line was averaged from 20 frames, temporal end. Negative correlation
without eye-tracking mode. with age (r = -0.62, P < .0001).
OCT = optical coherence tomography; AMD = age-related macular degeneration; PED = pigment epithelium detachment; D = diopter; CSC = central serous chorioretinopathy; LP = laser photocoagulation; PDT =
photodynamic therapy; CR = coefficient of repeatability.
a
Choroidal thickness defined as distance between outer border of retinal pigment epithelium to inner scleral border.
Heidelberg Spectralis is manufactured by Heidelberg Engineering, Heidelberg, Germany. Cirrus HD-OCT is manufactured by Carl Zeiss Meditec, Inc., Dublin, CA.

Copyright SLACK Incorporated

 i m a g i n g

20. Rahman W, Chen FK, Yeoh J, Patel P, Tufail A, Da Cruz L. Repeat-

able SD-OCTs. Its operation is simple, it is non-inva- ability of manual subfoveal choroidal thickness measurements in
sive, it provides an in vivo account of the choroid, and healthy subjects using the technique of enhanced depth imaging opti-
it has high repeatability. With this technique, choroidal cal coherence tomography [published online ahead of print Novem-
ber 18, 2010]. Invest Ophthalmol Vis Sci.
changes in some retinal and choroidal diseases were re- 21. Margolis R, Spaide RF. A pilot study of enhanced depth imaging opti-
vealed. New insights into the pathogenesis and treat- cal coherence tomography of the choroid in normal eyes. Am J Oph-
thalmol. 2009;147:811-815.
ment response were reported, which may light the path 22. Ramrattan RS, van der Schaft TL, Mooy CM, de Bruijn WC, Mulder
for future research. To date, published data regarding PG, de Jong PT. Morphometric analysis of Bruchs membrane, the
choriocapillaris, and the choroid in aging. Invest Ophthalmol Vis Sci.
EDI-OCT are still limited. There is a demand for more 1994;35:2857-2864.
work to be done. 23. Manjunath V, Taha M, Fujimoto JG, Duker JS. Choroidal thickness
in normal eyes measured using Cirrus HD optical coherence tomog-
raphy. Am J Ophthalmol. 2010;150:325-329.e1.
REFERENCES 24. Fujiwara T, Imamura Y, Margolis R, Slakter JS, Spaide RF. Enhanced
depth imaging optical coherence tomography of the choroid in highly
1. Huang D, Swanson EA, Lin CP, et al. Optical coherence tomography. myopic eyes. Am J Ophthalmol. 2009;148:445-450.
Science. 1991;254:1178-1181. 25. Macular Photocoagulation Study Group. Subfoveal neovascular le-
2. Hee MR, Izatt JA, Swanson EA, et al. Optical coherence tomography sions in age-related macular degeneration: guidelines for evaluation
of the human retina. Arch Ophthalmol. 1995;113:325-332. and treatment in the macular photocoagulation study. Arch Ophthal-
3. Puliafito CA, Hee MR, Lin CP, et al. Imaging of macular diseases with mol. 1991;109:1242-1257.
optical coherence tomography. Ophthalmology. 1995;102:217-229. 26. Bird AC, Marshall J. Retinal pigment epithelial detachments in the
4. Hee MR, Puliafito CA, Wong C, et al. Quantitative assessment of elderly. Trans Ophthalmol Soc U K. 1986;105:674-682.
macular edema with optical coherence tomography. Arch Ophthalmol. 27. Casswell AG, Kohen D, Bird AC. Retinal pigment epithelial detach-
1995;113:1019-1029. ments in the elderly: classification and outcome. Br J Ophthalmol.
5. Massin P, Girach A, Erginay A, Gaudric A. Optical coherence to- 1985;69:397-403.
mography: a key to the future management of patients with diabetic 28. Kuhn D, Meunier I, Soubrane G, Coscas G. Imaging of chorioretinal
macular oedema. Acta Ophthalmol Scand. 2006;84;466-474. anastomoses in vascularized retinal pigment epithelium detachments.
6. Panozzo G, Gusson E, Parolini B, Mercanti A. Role of OCT in the di- Arch Ophthalmol. 1995;113:1392-1398.
agnosis and follow-up of diabetic macular edema. Semin Ophthalmol. 29. Slakter JS, Yannuzzi LA, Schneider U, et al. Retinal choroidal anasto-
2003;18:74-81. moses and occult choroidal neovascularization in age-related macular
7. Leung CK, Cheung CY, Weinreb RN, et al. Evaluation of retinal degeneration. Ophthalmology. 2000;107:742-753.
nerve fiber layer progression in glaucoma: a study on optical coher- 30. Yannuzzi LA, Negro S, Iida T, et al. Retinal angiomatous prolifera-
ence tomography guided progression analysis. Invest Ophthalmol Vis tion in age-related macular degeneration. Retina. 2001;21:416-434.
Sci. 2010;51:217-222. 31. Gass JD, Agarwal A, Lavina AM, Tawansy KA. Focal inner retinal
8. Cioffi GA, Granstam E, Alm A. Ocular circulation. In: Kaufman PL, hemorrhages in patients with drusen: an early sign of occult cho-
Alm A, eds. Aders Physiology of the Eye: Clinical Application, 10th ed. roidal neovascularization and chorioretinal anastomosis. Retina.
St. Louis: Mosby; 2003:747-784. 2003;23:741-751.
9. Muller-Delp JM. Aging-induced adaptations of microvascular reactiv- 32. Gass JD. Pathogenesis of disciform detachment of the neuroepithe-
ity. Microcirculation. 2006;13:301-314. lium. Am J Ophthalmol. 1967;63(suppl):1-139.
10. Yildiz O. Vascular smooth muscle and endothelial functions in aging. 33. Spaide RF, Goldbaum M, Wong DW, Tang KC, Iida T. Serous de-
Ann N Y Acad Sci. 2007;1100:353-360. tachment of the retina. Retina. 2003;23:820-846.
11. Farkas E, Luiten PG. Cerebral microvascular pathology in aging and 34. Hayashi K, Hasegawa Y, Tokoro T. Indocyanine green angiography of
Alzheimers disease. Prog Neurobiol. 2001;64:575-611. central serous chorioretinopathy. Int Ophthalmol. 1986;9:37-41.
12. Futrakul N, Futrakul P. Renal microvascular disease in an aging popu- 35. Guyer DR, Yannuzzi LA, Slakter JS, Sorenson JA, Ho A, Orlock D.
lation: a reversible process? Ren Fail. 2008;30:353-356. Digital indocyanine green videoangiography of central serous chorio-
13. Spaide RF, Koizumi H, Pozonni MC. Enhanced depth imaging retinopathy. Arch Ophthalmol. 1994;112:1057-1062.
spectral-domain optical coherence tomography. Am J Ophthalmol. 36. Spaide RF, Hall L, Haas A, et al. Indocyanine green videoangiogra-
2008;146:496-500. phy of older patients with central serous chorioretinopathy. Retina.
14. Povazay B, Hermann B, Unterhuber A, et al. Three-dimensional 1996;16:203-213.
optical coherence tomography at 1050 nm versus 800 nm in reti- 37. Spaide RF, Campeas L, Haas A, et al. Central serous chorioretinopathy
nal pathologies: enhanced performance and choroidal penetration in in younger and older adults. Ophthalmology. 1996;103:2070-2079.
cataract patients. J Biomed Opt. 2007;12:041211. 38. Shiraki K, Moriwaki M, Matsumoto M, Yanagihara N, Yasunari
15. Spaide RF. Enhanced depth imaging optical coherence tomography of T, Miki T. Long-term follow-up of severe central serous chorioreti-
retinal pigment epithelial detachment in age-related macular degen- nopathy using indocyanine green angiography. Int Ophthalmol. 1997-
eration. Am J Ophthalmol. 2009;147:644-652. 1998;21:245-253.
16. Ikuno Y, Kawaguchi K, Nouchi T, Yasuno Y. Choroidal thickness in 39. Iida T, Kishi S, Hagimura N, Shimizu K. Persistent and bilateral
healthy Japanese subjects. Invest Ophthalmol Vis Sci. 2010;51:2173- choroidal vascular abnormalities in central serous chorioretinopathy.
2176. Retina. 1999;19:508-512.
17. An L, Wang RK. Use of a scanner to modulate spatial interferograms 40. Prnte C, Flammer J. Choroidal capillary and venous congestion in
for in vivo full-range Fourier-domain optical coherence tomography. central serous chorioretinopathy. Am J Ophthalmol. 1996;121:26-34.
Opt Lett. 2007;32:3423-3425. 41. Giovannini A, Scassellati-Sforzolini B, DAltobrando E, Mariotti C,
18. Wang RK. Fourier-domain optical coherence tomography achieves Rutili T, Tittarelli R. Choroidal findings in the course of idiopathic se-
full range complex imaging in vivo by introducing a carrier frequency rous pigment epithelium detachment detected by indocyanine green
during scanning. Phys Med Biol. 2007;52:5897-5907. videoangiography. Retina. 1997;17:286-293.
19. Vakoc BJ, Yun SH, Tearney GJ, Bouma BE. Elimination of depth 42. Tittl M, Polska E, Kircher K, et al. Topical fundus pulsation measure-
degeneracy in optical frequency-domain imaging through polariza- ment in patients with active central serous chorioretinopathy. Arch
tion-based optical demodulation. Opt Lett. 2006;31:362-364. Ophthalmol. 2003;121:975-978.

Ophthalmic Surgery, Lasers & Imaging Vol. 42, No. 4 (Suppl), 2011 S83
 i m a g i n g

43. Tittl M, Maar N, Polska E, Weigert G, Stur M, Schmetterer L. Cho- 2008;115:1756-1765.

roidal hemodynamic changes during isometric exercise in patients 55. Schmidt-Erfurth U, Hasan T, Gragoudas E, Michaud N, Flotte TJ,
with inactive central serous chorioretinopathy. Invest Ophthalmol Vis Birngruber R. Vascular targeting in photodynamic occlusion of sub-
Sci. 2005;46:4717-4721. retinal vessels. Ophthalmology. 1994;101:1953-1961.
44. Imamura Y, Fujiwara T, Margolis R, Spaide RF. Enhanced depth im- 56. Kramer M, Miller JW, Michaud N, et al. Liposomal benzoporphy-
aging optical coherence tomography of the choroid in central serous rin derivative verteporfin photodynamic therapy: selective treat-
chorioretinopathy. Retina. 2009;29:1469-1473. ment of choroidal neovascularization in monkeys. Ophthalmology.
45. Tewari HK, Gadia R, Kumar D, Venkatesh P, Garg SP. Sympathetic-para- 1996;103:427-438.
sympathetic activity and reactivity in central serous chorioretinopathy: a 57. Schmidt-Erfurth U, Laqua H, Schltzer-Schrehard U, Viestenz A,
case-control study. Invest Ophthalmol Vis Sci. 2006;47:3474-3478. Naumann GO. Histopathological changes following photodynamic
46. Michael JC, Pak J, Pulido J, de Venecia G. Central serous chorioreti- therapy in human eyes. Arch Ophthalmol. 2002;120:835-844.
nopathy associated with administration of sympathomimetic agents. 58. Schltzer-Schrehardt U, Viestenz A, Naumann GO, Laqua H, Mi-
Am J Ophthalmol. 2003;136:182-185. chels S, Schmidt-Erfurth U. Dose-related structural effects of photo-
47. Tittl MK, Spaide RF, Wong D, et al. Systemic findings associated with dynamic therapy on choroidal and retinal structures of human eyes.
central serous chorioretinopathy. Am J Ophthalmol. 1999;128:63-68. Graefes Arch Clin Exp Ophthalmol. 2002;240:748-757.
48. Karadimas P, Bouzas EA. Glucocorticoid use represents a risk factor 59. Isola V, Pece A, Parodi MB. Choroidal ischemia after photodynamic
for central serous chorioretinopathy: a prospective, case-control study. therapy with verteporfin for choroidal neovascularization. Am J Oph-
Graefes Arch Clin Exp Ophthalmol. 2004;242:800-802. thalmol. 2006;142:680-683.
49. Haimovici R, Koh S, Gagnon DR, Lehrfeld T, Wellik S, Central Se- 60. Piccolino FC, Borgia L. Central serous chorioretinopathy and indo-
rous Chorioretinopathy Case-Control Study Group. Risk factors for cyanine green angiography. Retina. 1994;14:231-242.
central serous chorioretinopathy: a case-control study. Ophthalmology. 61. Iida T, Kishi S, Hagimura N, Shimizu K. Persistent and bilateral
2004;111:244-249. choroidal vascular abnormalities in central serous chorioretinopathy.
50. Maruko I, Iida T, Sugano Y, et al. Subfoveal choroidal thickness af- Retina. 1999;19:508-512.
ter treatment of central serous chorioretinopathy. Ophthalmology. 62. Yeoh J, Rahman W, Chen F, et al. Choroidal imaging in inherited
2010;117:1792-1799. retinal disease using the technique of enhanced depth imaging op-
51. Yannuzzi LA, Slakter JS, Gross NE, et al. Indocyanine green angiog- tical coherence tomography. Graefes Arch Clin Exp Ophthalmol.
raphy-guided photodynamic therapy for treatment of chronic central 2010;248:1719-1728.
serous chorioretinopathy: a pilot study. Retina. 2003;23:288-298. 63. Coleman DJ, Silverman RH, Chabi A, et al. High-resolution ultrasonic
52. Cardillo Piccolino F, Eandi CM, Ventre L, Rigault de la Longrais RC, imaging of the posterior segment. Ophthalmology. 2004;111:1344-1351.
Grignolo FM. Photodynamic therapy for chronic central serous cho- 64. Weber AL, Mafee MF. Evaluation of the globe using computed tomogra-
rioretinopathy. Retina. 2003;23:752-763. phy and magnetic resonance imaging. Isr J Med Sci. 1992;28:145-152.
53. Chan WM, Lai TY, Lai RY, Tang EW, Liu DT, Lam DS. Safety en- 65. Sarks SH. Aging and degeneration in the macular region: a clinico-
hanced photodynamic therapy for chronic central serous chorioretinop- pathological study. Br J Ophthalmol. 1976;60:324-341.
athy: one-year results of a prospective study. Retina. 2008;28:85-93. 66. Yeoh J, Rahman W, Chen F, et al. Choroidal imaging in inherited
54. Chan WM, Lai TY, Lai RY, Liu DT, Lam DS. Half-dose vertepor- retinal disease using the technique of enhanced depth imaging op-
fin photodynamic therapy for acute central serous chorioretinopathy: tical coherence tomography. Graefes Arch Clin Exp Ophthalmol.
one year results of a randomized controlled trial. Ophthalmology. 2010;248:1719-1728.

S84 Copyright SLACK Incorporated