use of a combined slit-lamp sd-oct to obtain anterior and posterior segment images in selected...

Use of a combined slit-lamp SD-OCT to obtain anterior andposterior segment images in selected animal species

Serge G. Rosolen,*,†,‡ Mei-Lyn K. Riviere,* Sylvie Lavillegrand,* Barbara Gautier,§Serge Picaud†,‡,

–,**,†† and Jean-Francois LeGargasson†,‡‡

*Clinique Veterinaire, Asnieres, France; †INSERM, U968, Institut de la Vision, Paris, F-75012, France; ‡UPMC Univ Paris 06, UMR_S968, Institut de

la Vision, Paris, F-75012, France; §Clinique Veterinaire, Freneuse, France; –CNRS, UMR 7210, Institut de la Vision, Paris, F-75012, France; **Centre

Hospitalier National d’Ophtalmologie des Quinze-Vingts, Paris, France; ††Fondation Ophtalmologique Adolphe de Rothschild, Paris, France; and ‡‡UDD

Paris 7, Paris, France

Address communications to:

S. G. Rosolen

Tel.: +33 1 47 33 08 95

Fax: +33 1 47 33 30 37

e-mail: [email protected]

AbstractObjective To obtain images of anterior and posterior segments of the eye using a slit-lamp (SL)/spectral domain (SD) optical coherence tomography (OCT) integrated

system designed for the human eye, in the cat, dog, minipig and monkey.Animals studied One healthy adult monkey, one healthy adult minipig, one healthy

adult dog, one healthy adult cat, and three cats and four dogs affected by corneal orretinal diseases.

Procedure A SL SCAN-1 SD-OCT, which is a slit-lamp SL-D7 that contains anintegrated OCT module and a fundus viewer, was used to generate OCT images

(512–2048), while simultaneously taking ‘en-face’ slit-lamp images (efSL). OCTimages were obtained under sedation or anesthesia. These images were compared tohistological retinal sections obtained from a monkey, a minipig, a dog, and a cat.

Results ‘en-face’ slit-lamp images and OCT images of the ocular tissues were obtainedallowing for the identification of different corneal and retinal layers in all animal

species. Measurements of the total retinal thickness (TRT) from the inner limitingmembrane to the retinal pigment epithelium were performed in various regions

throughout the retina. Reduction in TRT was consistent with clinical features ofretinal degeneration identified in dogs and cats.

Conclusion This noninvasive procedure is useful for both experimental and clinicalassessments of ocular tissue damage. Images of anterior and posterior segments are

readily obtained under routine clinical conditions. Future studies are warranted toestablish normal OCT data in our patients with this new instrument.

Key Words: anterior segment, cat, dog, minipig, monkey, optical coherence

tomography, posterior segment

INTRODUCTION

Optical coherence tomography (OCT) is a noninvasive,optical, diagnostic technique that provides depth resolutionof images of tissues up to a few micrometers and depths upto several millimeters in both transparent and nontranspar-ent biological tissues.1–3 OCT is analogous to B-mode ultra-sonography using optical tissue reflectivity rather thanacoustic tissue reflectivity. Since introduced in 1991, thistechnique has been extensively applied in many bio-medicalareas. OCT is particularly attractive for ophthalmic imag-ing, providing optical sections of the anterior segment and

the retina, thereby allowing for the diagnosis of ocular dis-eases.4,5

Briefly, the principle of OCT, based on a Michelson inter-ferometer, is to detect photons that are backscattered by thetissue with respect to the coherent length of the source. Overthe last decade, the SD-OCT has become an importantinstrument for eye examinations in patients by ophthalmolo-gists.6,7 OCT is currently used for both anterior8,9 and pos-terior segment examinations.10,11 Anterior segment OCTprovides structural information of the cornea and anteriorchamber without directly contacting the eye, thus offeringan ease of image acquisition and a considerable advantage

� 2012 American College of Veterinary Ophthalmologists

Veterinary Ophthalmology (2012) 1–11 DOI:10.1111/j.1463-5224.2012.01037.x

over ultrasound biomicroscopy (UBM); however, in contrastto the latter, it cannot be used to view deep structures suchas the ciliary body, for example. In addition, OCT offers ahigher axial resolution (5–10 lm) of the anterior segmentthan UBM (25 lm).12 Posterior segment OCT providesstructural information of the retina in a noninvasive manner,which is important for the diagnosis and follow-up of vari-ous blinding pathologies, including age-related maculardegeneration (AMD), diabetic retinopathy, glaucoma, andretinitis pigmentosa.

The high cost of available SD-OCT instruments has lim-ited its animal application to research studies in mice13,14

rats,15–17 rabbits,18,19 cats,20,21 dogs,22,23 pigs,24,25 and non-human primates.26,27 Recently, however, a more economicalOCT system integrated within a slit lamp was released forhuman clinical investigations.

Given the potential usefulness of the SD-OCT in veteri-nary ophthalmology clinics as well as the potential toupgrade from a slit lamp, we investigated whether this newsystem could provide appropriate imaging of the anteriorand posterior eye segments in selected animal species.

MATERIALS AND METHODS

The protocol adhered to the Association for Research inVision and Ophthalmology Statement for the Use ofAnimals in Ophthalmic and Vision Research. Dogs and catsaffected with ocular disorders were examined at the Veteri-nary Eye Clinic (Serge G Rosolen) after obtaining informedconsent from the owners.

AnimalsOCT data were obtained from a total of 11 animals, includ-ing four that were normal and were used for research andseven that were presented for evaluation of clinical signssuggestive of ocular pathology (see Table 1). One young

healthy adult male cynomolgus monkey (monkey #1),one young healthy adult female minipig (minipig #1),one young healthy neutered adult male shorthair cat (cat#1), and one young healthy adult male beagle dog (dog#1) were included in the study. One dog and two catsrevealed corneal diseases. More specifically, one 9-year-old male Shih Tzu presented with corneal mineraldeposits, one 5-year-old neutered female shorthairdomestic cat presented with chronic keratitis, and one4-year-old male Persian cat presented with cornealsequestrum. A total of three dogs and one cat revealedretinal diseases in the absence of any optics changes (i.e.,cataract, vitreal degeneration, for example) detected withthe slit-lamp examination. These included a 9-year-oldmale Yorkshire terrier with no visual defect, but whopresented ophthalmoscopically some areas of hyper-reflectivity, a 7-year-old male miniature poodle present-ing with clinical visual defects in a mesopic environment,which was associated with hyper-reflectivity of the fun-dus, a blind 8-year-old male miniature poodle with signsof retinal atrophy, and finally a blind 8-year-old femalePersian cat with signs of retinal atrophy. None of theanimals used for OCT imaging was killed in this study.However, to compare OCT images with eye tissue his-tology, histological sections were performed on the eyesof animals euthanized for reasons unrelated to this study.Histological sections of normal cornea and normal retina(fixed in Davidson’s solution, embedded in paraffin andstained with hematoxylin-eosin) were obtained fromcomparable regions to those scanned in one young adultcynomolgus monkey (monkey #2), one young adult mini-pig (minipig #2), one young adult domestic cat (cat #2),and one young adult beagle dog (dog #2).

With the exception of the cynomolgus monkey thatwere anesthetized with a mixture of ketamine/medetomi-dine (0.1 mg/kg for medetomidine; 5 mg/kg for ketamine

Table 1. Dogs and cats signalment information and clinical features and their related Figures are presented in the table. None of the eyes presented

with an albinotic or subalbinotic fundus

Animal number Figure number Clinical signalment Clinical features

Anterior segmentDog #1 2a 1-year-old male Beagle dog NormalCat #1 2b 1-year-old male shorthair cat NormalDog #3 4a 9-year-old male Shih Tzu Corneal mineral depositCat #3 4b 5-year-old neutered female shorthair

domestic catKeratitis

Cat #4 4c 4-year-old male Persian cat Corneal sequestrumPosterior segment

Cat #1 6a-c 1-year-old male shorthair cat NormalDog #1 7a1–4 1-year-old male Beagle dog NormalDog #4 8b 9-year-old Yokshire No visual defect, no ophthalmic opacities,

slight hyper-reflectivity around the papilla,Dog #5 8c 7-year-old miniature Poodle Visual defect in mesopic environment, no ophthalmic

opacities, vessels tortuosity and many areas withhyper-reflectivity in the early periphery

Cat #5 8d 8-year-old Persian cat Blind with signs of retinal atrophyDog #6 8e 8-year-old miniature Poodle Blind with signs of retinal atrophy

2 r o s o l e n E T A L .

� 2012 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–11

hydrochloride), all animals were sedated with a single intra-muscular injection of medetomidine (Domitor�; Pfizer,Orsay, France) at a dose of 0.1 mg/kg for cats and minipigsand at dose of 0.03 mg/kg for dogs. A complete physicalexamination was carried out in all animals to ensure thatthere was no contraindication to anesthesia or sedation.Optical coherence tomography was carried out in a dimlylight room, and pupils were dilated using 1% tropicamideeye drops (Mydriaticum�; Thea, Rueil, France) to standard-ize the procedure. Although SD-OCT facilitates the rapidA-scan collection, the quality of SD-OCT images canalso be altered by eye movements (such as nystagmus orParkinson’s disease, for example) as well as by noncoopera-tive patients. The eye was stabilized during the process ofimage acquisition by means of an atraumatic conjunctivalclip (SIEM-Biomedicale, Nımes, France), as previouslydescribed.28 This facilitated manipulation of the eye, assuredproper alignment in front of the scanning beam to obtainmeasurements perpendicular to the corneal surface as wellas orientation to allow for examination of specific retinalareas (i.e., papilla, area with or without the tapetum luci-dum). While globe fixation indeed facilitates OCT imageacquisition, it is not a prerequisite given that animals aresedated or anesthetized.

OCT imagesOCT images were obtained with a commercial unit (Top-con SL SCAN-1; Topcon, Clichy, France) in which theSD-OCT is integrated with a SL-D7 slit lamp (light source,super luminescent light emitting diode of 840 nm; axial res-olution, 8–9 lm; lateral resolution, <20 lm; pixel imagedepth, 512–2048; scanning speed, 5000 A-scans/s and oneB-scan (512 A-scans) in 0.1 s; scanning range, 2–12 mmposterior and 8–12 mm anterior; scan width, 2.3 mm; scanpatterns: horizontal line, vertical line, cross, raster, grid,radial, circle) integrated in a slit-lamp SL-D7 (five differentmagnifications: 6·, 10·, 16·, 25·, and 40·). The maximumscanning time is limited to 1 s to reduce movement artifacts.The system also contains a fundus viewer (field of view >30�)coupled to a digital camera (Topcon DC3) allowing for thevisualization, recording and capture of en-face images whilesimultaneously generating the cross-sectional images takenby the OCT system. Figure 1 illustrates the setup of the ani-mal examination on sedated cat #1. Anterior segment imageswere obtained without any additional optical components,whereas fundus images required an additional optical device(Volk 78D Classic lens) fixed on a specific lens mount. Theworking distance between the lens and the cornea was<2 cm, and corneal hydration was maintained with eye drops(one drop at four times/min) (BSS; Alcon, Rueil, France)that were applied by an assistant. Eye structures assessed(anterior segment or posterior segment) and their relatedtomograms were visualized on a monitor by TOPCON’Ssoftware package IMAGE net i-base. Once the focus obtainedwith the SL was correct, SD-OCT scanning acquisition wasinitiated by depressing a joystick button.

For imaging of the anterior segment, the setup parametersrecommended by the manufacturer for optimized imageswere followed: scanning 10 mm with 10· slit-lamp magnifi-cation whatever the scan patterns used. The softwareprovides the possibility of manually measuring the thicknessof tissues.

For imaging the retina, the setup parameters recom-mended by the manufacturer for optimized images were alsofollowed: scanning 6 mm using 16· slit-lamp magnificationwhatever the scan patterns used.

For retinal image analysis, the manufacturer’s softwareprovides an algorithm that generates automatically twolines: delineating the inner limiting membrane (ILM) andthe retinal pigment epithelium (RPE). The software alsoallows for manual correction of these delineations. Thedistance between these lines is designated as the total reti-nal thickness (TRT) by the manufacturer. Because the sys-tem was calibrated for human examination, the position ofthese lines was visually controlled prior to TRT measure-ments and manually corrected when necessary. A softwarecaliper tool offers the option to manually measure thethickness of select regions and distance between the delin-eated lines. Table 2 summarizes the parameters usedaccording to the eye structure being imaged as well as theanimal species.

RESULTS

Table 1 summarizes signalment information and clinicalfeatures of dogs and cats evaluated.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

Figure 1. SD-OCT examination in normal cat (cat #1). The sedated

cat is positioned in sternal recumbency, and the manipulation of the eye

is made with a clip (a). Corneal hydration is continued throughout the

complete examination (b). The Topcon SL SCAN-1 comprises a table

SL-D7 slit lamp (c) including a scan unit (d), a digital camera DC-3 (e),

and a fundus viewer (f). An additional fixed optical device (g) (78 D

Classic Volk lens) is used to examine the posterior segment. When the

SL SCAN-1 begins, scans are acquired, while a live view of the fundus

is simultaneously observed on a screen (h). Images are captured by

pressing a joystick button (i).

u s e o f a c o m b i n e d s l i t - l a m p s d - o c t 3


Corneal SD-OCT imagesCentral corneal SD-OCT images obtained from the normaldog, cat, and minipig are presented in Fig. 2a–c, respec-tively. A photomicrograph illustrating the histology of thenormal canine cornea (dog #2) is presented in Fig. 2d forcomparison. In all normal corneal SD-OCT images, fromthe top (exterior surface) to bottom (inner surface), alternat-ing dark-light signal bands represent distinct anatomic cor-neal layers. As indicated in Fig. 2a–c, a thin light-dark bandwas observed, which was attributable to the tear film/cornealepithelium complex (Tf/Ep), a large band was attributableto the corneal stroma (St), and a thin triple band attributableto the Descemet’s membrane/corneal endothelium (Dm/En) complex and were all clearly identifiable and could bemeasured manually.

In dog #3 (Fig. 3a,a1,a2), corneal mineral deposits wereclearly seen as dense white opacities on the en-face slit-lamp(efSL) images. Related SD-OCT images depicted altera-tions in the corneal surface, with irregularities in the epithe-lium, and highly reflective areas (bright signals) within thestroma, presumably corresponding to the mineral deposits.The triple band attributable to the Dm/En complex was notclearly identifiable as in the normal dog (dog #1).

In cat #3 with a history of keratitis (Fig. 3b,b1,b2), SD-OCT images demonstrated obvious corneal surface contourirregularity, lack of visualization of the Dm/En complextriple band, and apparent intra-stromal cavities.

In cat #4 with a corneal sequestrum (Fig. 3c,c1,c2), cor-neal SD-OCT images illustrate corneal surface alteration,with rupture of the corneal epithelium and apparent separa-tion of the sequestrum from the surrounding stroma. Man-ual measurements of the sequestrum indicated that it was

approximately 300 lm thick, although precise depth of cor-neal stromal involvement could not be determined becauseof drop-out of the optical signal posterior to this denselyopaque structure.

Posterior segment SD-OCT imagesRetinal SD-OCT images obtained from the normal monkey(centered on the macula of monkey #1), the normal minipig(minipig #1), the normal cat (cat #1), and the normal dog(dog #1) are presented in Figs 4b,5b,6a2–c2, and 7a, respec-tively. Photomicrographs illustrating the histology of thenormal retina of monkey (monkey #2), minipig (minipig #2),and cat (cat #2) are presented in Figs 4c,5d, and 6a3–c3 forcomparison.

In the normal monkey (monkey #1), the foveal depressionis clearly identifiable in both ‘in vivo’ (Fig. 4b, arrow) andhistological (Fig. 4c, arrow) images. From the proximal (top)retina to the distal (bottom) retina, alternate dark-lightbands of signals in SD-OCT tomograms directly correlatewith the retinal layers. As in human OCT images, one canattribute the different layering in the nonhuman primateOCT image to the different retinal layers. As suggested inFig. 4d, the proximal thin band can be attributed to ganglioncell axons, also called the nerve fiber layer (NFL), whereaslarger bands attributable to the ganglion cell layer (GCL),the inner plexiform layer (IPL), the inner layer (INL), and tothe outer nuclear layer (ONL) were also clearly identifiable.While photoreceptor inner segments (IS) and photoreceptorouter segments (OS) were clearly identifiable in histologicalsections, only the interface between both was visible in SD-OCT tomograms. A triple band appearance described in theliterature5 and commonly attributable to the interface of

Table 2. Parameters for acquisition of OCT images from anterior (scanning speed: 5000 a-scans/s and one B-scan [512 a-scans] in 0.1 s;

length [10 mm]; slit-lamp magnification [10·]) (A) and posterior(scanning speed: 5000 A-scans/s and one B-scan [512 a-scans] in 0.1 s;

length [6 mm]; slit-lamp magnification [16·]) (B) segments are indicated in the table

Animal Figure number B-scan pattern number (format) Sampling depth (pixels)

(A)Normal conditions

Minipig #1 3c 6 (radial) 512Cat #1 3b 6 (radial) 512Dog #1 3a 6 (radial) 512

Pathological conditionsShih Tzu (Dog #3) 4a 6 (radial) 512Domestic cat (Cat #3) 4b 8 (grid) 5124 year old Persian cat (Cat #4) 4c 8 (grid) 512

(B)Normal conditions

Monkey #1 5 1 1024Minipig #1 6 8 (grid) 512Cat #1 7 6 (radial) 512Dog #1 8a 6 (radial) 512

Pathological conditions8-year-old Persian cat (Cat #5) 8d 6 (radial) 5129-year-old Yorkshire (Dog #4) 8b 6 (radial) 5127-year-old miniature Poodle (Dog #5) 8c 6 (radial) 5128-year-old miniature Poodle (Dog #6) 8e 6 (radial) 512



photoreceptor inner segments (IS)/photoreceptor outersegment (OS)/RPE–choriocapillaris (CC)/choroid (CH)was clearly identifiable on SD-OCT tomograms.

In the normal minipig (minipig #1), the retina has a morehomogenous appearance, where from the proximal retina(top) to the distal retina (bottom), a large white band attrib-utable to the GCL and a large dark band attributable to theONL were identified. It is of note that the triple bandobserved in the macular region of the monkey was not iden-tifiable in the minipig, a feature previously described in theliterature.24 Therefore, a white band attributable to OS/IS/RPE/CC clearly separated the previous layers from an alter-nate dark-white segment attributable to the CH, includinglarge dark spaces that are presumably attributable to choroi-dal vessels. In Fig. 5b, while the NFL is not clearly identifi-able in the proximal retina, three dark transverse bands(Fig. 5b, arrows) are clearly seen to extend beyond protuber-ance in the NFL that correspond to the location of retinalblood vessels.

In the normal cat (cat #1), SD-OCT images were exam-ined in three different areas (at the papilla in Fig. 6a, nearbythe area centralis in Fig. 6b and in Fig. 6c in a region lackingthe tapetum lucidum, located at a distance from the papillaequivalent to that performed in Fig. 6b) of the right eye of ayoung healthy adult male neutered shorthair cat (cat #1). InFig. 6a2,b2,c2, selected B-scans include the visualization oftwo green lines representing the inner limiting membrane(ILM on top) and the retinal pigment epithelium (RPE atbottom), respectively. These lines were automatically gener-ated by application of a software algorithm. Qualitative

analysis of the B-scan performed in the region of the papilla(Fig. 6a2) depicted a trough at the level of the optic nervehead and a large heterogeneous white-prominent/dark bandpresumably attributable to NFL/GCL and another dark-prominent/white presumably attributable to IPL/INL/ONL layers in regions adjacent to the optic nerve head.Similar to the minipig, a triple band was not identifiable bycontrast to the macular area in the monkey. However, awhite band presumably attributable to the OS/IS/RPE/CCinterface was visible. Similar to minipigs, large dark spacespresumably attributable to choroidal vessels were visible inthe choroid (CH). In both regions with (Fig. 6b2) and with-out TL (Fig. 6c2) and from proximal (top) to distal (bottom),SD-OCT tomograms looked similar with a large dark-prominent band and a thinner white-prominent band. Simi-lar to the minipig, the cat retina has a more homogenousappearance than the monkey retina in SD-OCT images.This feature was previously reported in the literature.21

B-scan profiles enabled an automated quantitative evalua-tion of the retinal thickness.

Sampling values of the automated TRT are indicated inthe grid superimposed in the efSL eye fundus images and areillustrated in Fig. 6a1,b1,c1. In this example, TRT mean val-ues obtained with values displayed in Fig. 6a1 (n = 8), 6b1(n = 9), and 6c1 (n = 9) were 245.9 ± 22.8 lm in the periph-ery of the papilla, 196.9 ± 6.5 lm in the region adjacent tothe area centralis (with TL), and 155.6 ± 11.7 lm in theregion without TL, respectively.

SD-OCT posterior segment images obtained in dogs andcats that reveal retinal pathologies are presented in

EpSt

Tf

St

DmEn

1 mm

Ep

Tf

St

Dm/

1 mm 1 mm

(a) (d)

(f)(b)

x100

Figure 2. Anterior segment SD-OCT images obtained in central cornea in normal conditions (dog #1, cat #1, and minipig #1) according to the

parameters indicated in Table 1. Portions of original SD-OCT corneal images obtained from B-scans in dog #1, cat #1, and minipig #1 are presented

in a–c, respectively. Tear film (Tf), corneal epithelium (Ep), and corneal stroma (St), Descemet’s membrane (Dm) and corneal endothelium (En)

layers were identified and compared with a corneal section obtained from dog #2 (original magnification 100·) (Fig. 3d). In (a) (dog #1), manual

measurements of the Tf/Ep complex, stroma, and Dm/En complex were 75, 491, and 52 lm, respectively. In (b) (cat #1) and (c) (minipig #1), manual

measurements of total corneal thickness were 692/693 and 844/756 lm for the cat and minipig, respectively.



Fig. 7b–e. EfSL eye color fundus images, including thesuperposition of a grid that indicates the sampling of valuesof the total retinal thickness (TRT), are presented in line 1.The same black and white eye fundus images are presentedin line 2 to emphasize detail. B-scans selected among a radialpattern of six scans, including visualization of the ILM andRPE green lines along with their related profiles, are pre-sented in line 3 and line 4, respectively.

In dog #4 presenting with no visual defect, no ophthalmicopacities but slight hyper-reflectivity around the papilla(Fig. 7b1–b4) could be observed, and the positive wave inthe profile graph of the B-scan indicates a deviation that isattributed to a retinal vessel. The TRT mean value obtainedwith values (n = 9) displayed in Fig. 7b1 was 190.0 ±31.3 lm, a value very close to the TRT measured (n = 9;192.7 ± 30.2 lm) in the normal dog #1 (Fig. 7a1–a5).

In dog #5, presented with a visual defect in a mesopicenvironment, no ophthalmic opacities, vessel tortuosity inthe fundus as well as many areas of hyper-reflectivity in theearly periphery (Fig. 7c1–c4), the TRT mean value obtained

with values (n = 9) displayed in Fig. 7c1 was 92.3 ± 36.9 lm,indicating a decrease in approximately 50% in comparisonwith that measured in the normal dog #1 (Fig. 7a1–a4).These observations are consistent with an early stage of reti-nal atrophy, but we cannot relate vessel tortuosity with thisdisease.

In the blind cat #5 (Fig. 7d1–d4), presenting clinical signsof retinal atrophy without any ophthalmic opacity, it is ofinterest to note the thicker white band corresponding to theOS/IS/RPE/CC interface as well as a clear image of numer-ous vessels in the choroid. The TRT mean value obtainedwith values (n = 9) displayed in Fig. 7d1 was 83.0 ± 5.5 lmindicating a decrease in more than 50% in comparison withthe TRT mean value obtained in a similar area of cat #1(196.9 ± 6.5 lm) with values (n = 9) displayed in Fig. 6b2.

In the blind dog #6 (Fig. 7e1–e4), presenting clinical signsof retinal atrophy without any ophthalmic opacity, the B-scan depicted only a very thin white band and the TRTmean value obtained with values (n = 9) displayed inFig. 7e1 was 15.4 ± 13.8 lm. It is of note that the image of

x10

1

2

1

1 mm 1 mm

1

x10

x10

21 mm 1 mm

x10

1

22

1 mm1 mm

(a)

(b)

(c) (c1)

(b1)

(a1)

(c2)

(b2)

(a2)

Figure 3. Anterior segment SD-OCT images obtained from cats and dogs with corneal diseases. (a) Corneal mineral deposits in a 9-year-old male

Shih Tzu (dog #3). SD-OCT corneal images obtained from B-scans are presented in a1 for line 1 and in a2 for line 2. In a1, the corneal surface is

altered (arrowheads) and red arrows indicate an irregular limit between the stromal and corneal epithelium. In a2, arrows indicate stromal opacities.

(b) Keratitis in a 5-year-old neutered domestic shorthair female cat (cat #3). SD-OCT corneal images obtained from B-scans are presented in b1 for

line 1 and in b2 for line 2. In b1 and b2, arrowheads depict a significant alteration of the corneal surface, while in b1, red arrows indicate the rupture of

the Tf/Ep complex and white arrows indicate cavities that extend into the stroma. (c) Corneal sequestrum in a 4-year-old male Persian cat (cat #4).

SD-OCT corneal images obtained from B-scans are presented in b1 for line 1 and in b2 for line 2. In c1, arrowheads depict an important alteration of

the corneal surface while in c2, red arrows show the detachment of the sequestrum. Note that signal dropout is prominent on these images posterior

to the opacities.



the fundus is particularly poor, and the extreme retinal thin-ning measured could be artificially enhanced because of alack of signal strength.

DISCUSSION

The SL SCAN-1 designed for imaging human eyes enabledinvestigators to quickly obtain informative slit-lamp or fun-dus images while simultaneously acquiring SD-OCT scansof both anterior and posterior segments, in sedated or anes-thetized animals, enabling qualitative and quantitativeassessments of the cornea and/or retina. Given the fact thatcorneal desiccation and eye movements can degrade thequality of images, it is preferable to fix the globe and to pro-vide corneal hydration. Using these methods, two operatorsare required: one to provide corneal hydration and the otherto acquire the images. Care is necessary in order to properlyalign the structures to be studied perpendicularly to the cor-neal surface. SD-OCT image acquisition time is short, <1 sfor six radial B-scans. The complete procedure, includingsedation and eye positioning, lasts <10 min, which is a per-iod of time that is compatible with a routine procedureapplicable for clinical assessments. This quick procedure is

also applicable in a great number of animals for toxicologicalassessments. In normal animals of the species tested, imagesobtained with the use of the SL SCAN-1 in noncontact con-ditions provided structural information that was comparableto those described in the literature using other commercialSD-OCT. Other commercial SD-OCTs combine multi-modal equipment (scanning laser ophthalmoscopy, autoflu-orescence, fluorescein and ICG angiographies, ganglion cellcomplex and retinal nerve fiber layer analysis, microperime-try and eye tracking), thereby allowing one to gather addi-tional data, but also at the same time increasing the cost ofthe apparatus by a factor of two to fourfold compared to theSL SCAN-1.

For the anterior segment, the tear film/corneal epitheliumcomplex, the stroma, and the Descemet’s membrane/cornealendothelium complex were identifiable and their respectivethicknesses measurable. However, the goal of this work wasnot to establish statistical data of various corneal thicknesses.Nonetheless, measured values of illustrated examples areconsistent with previous studies showing that central corneaof the pig is thicker than that of carnivores.29,30 Cornealopacities such as corneal deposits or corneal sequestra opaci-ties degrade the image quality by masking deeper structures

16 x2001 x16 x2001 mm

GCL

NFL

GCL

IPL

OPL

OS/ISINL

GCL

INL

ONL RPECC

ONL

CH

IS RPEOS

CH

(a)

(d) (e)

(b) (c)

Figure 4. Posterior segment SD-OCT images obtained from normal monkey (monkey #1). In (a) an arrow identifies the macula, while a dotted line

indicates location of the B-scan. In (b) an arrow identifies the fovea in the OCT tomogram. In (c) an arrow identifies the fovea in histological section

obtained from monkey #2. In (d) and (e), portion of (b) and (c) are magnified, respectively. SD-OCT images clearly identify the nerve fiber layer

(NFL), ganglion cell body layer (GCL), inner plexiform layer (IPL), inner nuclear layer (INL), outer nuclear layer (ONL), and choroid (CH). A triple

band appearance is visualized in the SD-OCT images and has been attributed to the photoreceptor inner segment (IS)/photoreceptor outer segment

(OS)/RPE–chorio-capillaris (CC)/choroid (CH) interface. Two cross-retinal selective sections obtained manually with the use of the caliper (from

the NFL and OS/IS) measured 290 and 270 lm, respectively.



(such as the Dm/En complex, for example), a phenomenonprobably due to light absorption. In the case of keratitis,SD-OCT images revealed the heterogeneous nature of thetissue with the presence of intra-corneal cavities and anincrease in corneal thickness. Although the OCT technol-ogy did not allow visualization of infiltrating cells, the OCTimages suggest the presence of cellular infiltration that hadbeen described by others.31 A conventional slit-lamp exami-nation may not have revealed such cavities. It is of note thatthe accuracy of the corneal thickness measurement is limitedin the presence of corneal opacities owing to lack of opticalsignals which represent a major limitation of anteriorsegment OCT use. In animals, the SL SCAN-1 did not pro-vide access to deep anterior chamber structures such as theciliary body and irido-corneal angle and thus cannot replacethe use of UBM and gonioscopy. However, these differentexaminations could be carried out within the same sessionwhile animals are sedated. As corneal diseases represent amajor cause of visual impairment in animals,32,33 futurestudies will be aimed at investigating how this tool willenable us to document the evolution of corneal diseases inour patients.

For the posterior segment, images obtained with the SLSCAN-1 in the macular region of nonhuman primatesenable one to access structural information with imagequality comparable to that previously described in the liter-ature using other OCT systems.26,27,34,35 Similarly, the

images of retinal structures obtained from the pig are ofhigher quality with the SL SCAN-1 because of the applica-tion of SD-OCT as compared to previous studies.24,25 Indogs and cats, species in which a reflecting subretinal/cho-roidal structure (tapetum lucidum) exists, it is necessary toevaluate the impact of the latter on acquired images. What-ever the level of reflectivity (with the presence or absenceof tapetum lucidum), automated measurements of TRTwere taken and ILM and RPE layers were detected andsegmented by software on all images. Although the soft-ware offers the possibility to manually measure retinalthickness, it is preferable to use the automated measure-ments in order to avoid errors associated with possible mis-alignments with the manual use of a caliper. Our resultsconfirm that the retina was thicker near the papilla than inthe periphery. The slight difference of TRT measured inour animals between the tapetum lucidum regions andnon-tapetum lucidum regions might be due to its presenceand its thickness but also more probably due to the higherdensity of photoreceptor and retinal ganglion cells in thesuperior retinal regions corresponding to the area central-is.36 TRT measured values were comparable with thosemeasured in histological sections and described in the liter-ature.33,37 In regions without the tapetum lucidum, whenblood vessels were not clearly identifiable by funduscopy,SD-OCT B-scan profiles allowed for the identification oftheir presence (e.g., Fig. 6c2). In pathological conditions, a

x16 1 mm

NFL

GCL

INL

GCL

ONL

GCL

x400

ONLOSCHRPE

OS/IS + RPECC

RPE

CH

(a)

(c)

(b)

(d)

Figure 5. Posterior segment SD-OCT images obtained from normal minipig (minipig #1). In (a), a grid of eight cross-sections is superimposed in

an ‘en-face’ slit-lamp (efSL) fundus image. In (b) the green line indicates location of the B-scan. In (c), a magnified portion of (b) is compared in (d)

with a retinal section obtained in a similar region from minipig #2. From the proximal retina (top) to the distal retina (bottom), a large white band

attributable to the ganglion cell layer (GCL) and a large dark band attributable to outer nuclear layer (ONL) were identifiable. A white band attribut-

able to the OS/IS/RPE/CC clearly separated the above layers from an alternate dark-white segment attributable to the choroid (CH), including large

dark spaces presumably attributable to choroidal vessels. While the NFL was not clearly identifiable in the proximal retina, three dark transversal

bands (arrows in b), presumably attributable to retinal vessels could be observed. Two retinal cross-sections (from the NFL with and without OS/IS)

measured 264 and 284 lm, respectively, when measured manually.



reduction in retinal thickness was related to ophthalmo-scopically visible anomalies of the fundus. In our animals,this reduction in thickness was not attributed to changes inoptics of the eye because the slit-lamp examination did notdetect any anomalies such as cataract or vitreous opacities,but might therefore be attributed to retinal atrophy. How-ever, care must be taken in the interpretation of OCTimages given that opacities can degrade posterior segmentscan quality. Future studies are therefore necessary toinvestigate how this tool will enable us to document theevolution of retinal diseases in our patients.

In research, OCT images are already becoming veryinteresting for the documentation of retinal degenerationin rodents and in animals models that have larger eyes,such as chickens,38 pigs,21,24 cats,21 rabbits,18,19 dogs,23

and particularly nonhuman primates.26,27 OCT has thepotential to provide a better understanding of diseasedevelopment and progression in transgenic and othermodels of disease, which may translate to improvedclinical assessment and understanding of diseases, while

offering a platform for testing novel approaches totreatment.

In tested animal species, the SL SCAN-1 system builtfor human clinical use enables veterinarians or researchersto access structural information during a routine examina-tion. Further studies are warranted to establish how thefunduscopic examinations add to the simultaneousSD-OCT observation. Care must be taken, especially forposterior segment assessments, because the normal appear-ance of the fundus does not guarantee that the retinal tis-sue is structurally normal and use of the ERG would benecessary to assess the retinal function. The advantage ofcombining the SL SCAN-1 to study the structures of theeye and the acquisition of OCT images within the sameexamination could become an asset to the future develop-ment of veterinary ophthalmology. To our knowledge,there is currently no other device that integrates an OCTinto a slit lamp and that is commonly used for clinicalapplications at a lower cost than previously SD-OCTinstruments.

Retinal thickness map

B-scan n°4 2016 B scan n 4

NFLGCL

x20x16 1 mm

INL

ONLOS

B-scan n°6 CHTL

NFL

x400x16 1 mm

NFLGCL

INL

ONL

B-scan n°2

ONLOS

CH x400x16 1 mm

(a1)

(a2) (a3)

(b3)

(c3)

(b1)

(b2)

(b3)

(c1)

Figure 6. Posterior segment SD-OCT images obtained from normal cat (cat #1). A grid of six radial cross-sections is superimposed in ‘en-face’ slit-

lamp (efSL) fundus image in a region including the papilla (a1), an area in proximity to the area centralis (b1), and an area located in a region without

tapetum lucidum (c1) located at a distance from the papilla, equivalent to that performed in b1. A B-scan of cross-section #4 (from a1) is presented in

a2, a B-scan of cross-section #6 (from b1) is presented in b2, and a B-scan of cross-section #2 (from c1) is presented in c2. Two green lines, automati-

cally generated and representing the inner limiting membrane (ILM, above) and retinal pigment epithelium (RPE, below) are visible in a2, b2, and c2.

A magnified portion (red square) of a2, b2, and c2 are compared with a retinal section obtained from a similar region in cat #2 in a3, b3, and c3, respec-

tively. When measured manually with the help of the caliper, selective examples of the distance between ILM and RPE (from the NFL and OS/IS)

indicated 213 lm in a region near the papilla, 185 lm in a region with tapetum lucidum close to the area centralis and 151 lm in a region without

tapetum lucidum.



REFERENCES

1. Huang D, Swanson EA, Lin CP et al. Optical coherence tomog-

raphy. Science 1991; 254: 1178–1181.

2. Fujimoto JG, Brezinski ME, Tearney GJ et al. Optical biopsy and

imaging using optical coherence tomography. Nature Medicine1995; 1: 970–972.

3. Fercher AF. Optical coherence tomography. Journal of BiomedicalOptics 1996; 1: 157–173.

4. Puliafito CA, Hee MRE, Schuman JS et al. Optical Coherence Tomog-raphy of Ocular Disease. Slack Inc., Thorofare, NJ, 1995; p 374.

5. Drexler W, Fujimoto JG. State-of-the-art retinal optical coherence

tomography. Progress in Retina and Eye Research 2008; 27: 45–88.

6. Gabrielle ML, Wollstein G, Ishikawa H et al. Optical coherence

tomography: history, current status, and laboratory work. Investi-

gative Ophthalmology and Visual Science 2011; 52: 2425–2436.

7. Williams DR. Imaging single cells in the living retina. VisionResearch 2011; 51: 1379–1396.

8. Konstantopoulos A, Yadegarfar G, Fievez M et al. In vivo quantifi-

cation of bacterial keratitis with optical coherence tomography.

Investigative Ophthalmology and Visual Science 2011; 52: 1093–1097.

9. Vajzovic LM, Karp CL, Haft P et al. Ultra high-resolution ante-

rior segment optical coherence tomography in the evaluation of

the anterior corneal dystrophies and degenerations. Ophthalmology

2011; 118: 1291–1296.

10. Aref AA, Budenz DL. Spectral domain optical coherence tomog-

raphy in the diagnosis and management of glaucoma. OphthalmicSurgery and Lasers Imaging 2010; 41(Suppl): 15–17.

11. Lupo S, Grenga PL, Vingolo EM. Fourier-domain optical

coherence tomography and microperimetry findings in retinitis

pigmentosa. American Journal of Ophthalmology 2011; 151:

106–111.

12. Nolan W. Anterior segment imaging: ultrasound biomicroscopy

and anterior segment optical coherence tomography. Current

Opinion in Ophthalmology 2008; 19: 115–121.

13. Huber G, Beck SC, Grimm C et al. Spectral domain optical

coherence tomography in mouse model of retinal degeneration.


14. Gabriele ML, Ishikawa H, Schuman JS et al. Reproducibility of

spectral-domain optical coherence tomography total retinal thick-

ness measurements in mice. Investigative Ophthalmology and Visual

Science 2010; 51: 6519–6523.

15. Srinivasan VJ, Ko TH, Wojtkowski M et al. Non invasive

volumetric imaging and morphometry of the rodent retina with

high-speed, ultrahigh-resolution optical coherence tomography.


16. Nagata A, Higashide T, Ohkubo S et al. In vivo quantitative eval-

uation of rat retinal nerve fiber layer with optical coherence

tomography. Investigative Ophthalmology and Visual Science 2009;

50: 2809–2815.

1

2

x16 x16 x16 x16 x16

2

x16 x16 x16 x16 x16

3

mm1mm1mm1mm1mm1

4

Dog #1 Dog #4 Dog #5 Cat #5 Dog #6

(a) (b) (c) (d) (e)

Figure 7. Posterior segment SD-OCT images obtained in dogs and cats with retinal anomalies described in Table 2. Column (a) illustrates images

from the retina of a normal dog (dog #1), column (b) illustrates images from the retina of a 9-year-old male Yorkshire terrier (dog #4), column (c) illus-

trates images obtained from the retina of a 7-year-old male miniature poodle (dog #5), column (d) illustrates images obtained from the retina of an

8-year-old female Persian cat (cat #5), and column (e) illustrates images obtained from the retina of an 8-year-old male miniature poodle (dog #6).

‘en-face’ slit-lamp (EfSL) color fundus images, including the superposition of a grid of six radial cross-sections, are presented in line 1, while the same

efSL are presented in black and white to emphasize the details in line 2. B-scans selected from the grid of six radial cross-sections and their B-scan

profiles are presented in lines 3 and 4, respectively. As expected, measurements of total retinal thickness (TRT) of the selected B-scans and their

related profile revealed thinning of the retinal thickness according to the grade of atrophy. As expected, fundus images of retinal atrophy are correlated

with a significant thinning of the TRT; however, notably in the Yorkshire terrier, hyper-reflectivity does not appear to be correlated with a decrease

in the retinal thickness.



17. Bai Y, Xu J, Brahimi F et al. An agonistic TrkB mAb causes

sustained TrkB activation, delays RGC death, and protects the

retinal structure in optic nerve axotomy and in glaucoma. Investi-gative Ophthalmology and Visual Science 2010; 51: 4722–4751.

18. Hosseini K, Kholodnykh AI, Petrova IY et al. Monitoring of

rabbit cornea response to dehydration stress by optical coherence

tomography. Investigative Ophthalmology and Visual Science 2004;

45: 2555–2562.

19. Ameri H, Kim JG, Ratanapakorn T et al. Intravitreal and subreti-

nal injection of tissue plasminogen activator (tPA) in the treat-

ment of experimentally created retinal vein occlusion in rabbits.

Retina 2008; 28: 350–355.

20. Buhren J, Nagy L, Swanton JN et al. Optical effects of anti-

TGFbeta treatment after photorefractive keratectomy in a cat

model. Investigative Ophthalmology and Visual Science 2009; 50:

634–643.

21. Gekeler F, Gmeiner H, Volker M et al. Assessment of the poster-

ior segment of the cat eye by optical coherence tomography

(OCT). Veterinary Ophthalmology 2007; 10: 173–178.

22. Grozdanic SD, Matic M, Betts DM et al. Recovery of canine ret-

ina and optic nerve function after acute elevation of intraocular

pressure: implications for canine glaucoma treatment. Veterinary

Ophthalmology 2007; 10(suppl.1): 101–107.

23. Lheriteau E, Libeau L, Stieger K et al. The RPGRIP1-deficient

dog, a promising canine model for gene therapy. Molecular Vision2009; 15: 349–361.

24. Gloesmann M, Hermann B, Schubert C et al. Histologic correla-

tion of pig retina radial stratification with ultrahight-resolution

coherence tomography. Investigative Ophthalmology and Visual

Science 2003; 44: 1696–1703.

25. Gekeler F, Szurman P, Grisanti S et al. Compound subretinal

prostheses with extra-ocular parts designed for human trials: suc-

cessful long-term implantation in pigs. Graefe’s Archive for Clinical

and Experimental Ophthalmology 2007; 245: 230–241.

26. Schuman JS, Pedut-Kloizman T, Pakter H et al. Optical coher-

ence tomography and histologic measurements of nerve fiber

layer thickness in normal and glaucomatous monkey eyes. Investi-

gative Ophthalmology and Visual Science 2007; 48: 3645–3654.

27. Strouthidis NG, Grimm J, Williams GA et al. A comparison of

optic head morphology viewed by spectral domain optical coher-

ence tomography and by serial histology. Investigative Ophthalmol-ogy and Visual Science 2010; 51: 1464–1474.

28. Rosolen SG, Lamory B, Harms F et al. Cellular-resolution in vivo

imaging of the feline retina using adaptive optics: preliminary

results. Veterinary Ophthalmology 2010; 13: 369–376.

29. Samuelson DA. Ophthalmic anatomy. In: Veterinary Ophthalmol-

ogy, 4th edn. (ed. Gelatt KN) Blackwell publishing, Ames, 2007;

37–148.

30. Brunette I, Rosolen SG, Carrier M et al. Comparison of the pig

and feline models for full thickness corneal transplantation. Veter-

inary Ophthalmology 2011; 14: 365–377.

31. Prasse KW, Winston SM. Cytology and histopathology of feline

eosinophilic keratitis. Veterinary and Comparative Ophthalmology1996; 6: 74–81.

32. Gilger BC. Diseases and surgery of the canine cornea and sclera.

In: Veterinary Ophthalmology, 4th edn. (ed. Gelatt KN) Blackwell

publishing, Ames, 2007; 691–751.

33. Stiles J, Townsend WM. Feline ophthalmology. In: Veterinary

Ophthalmology, 4th edn. (ed. Gelatt KN) Blackwell publishing,

Ames, 2007; 1095–1164.

34. Toth CA, Narayan DG, Boppart SA et al. A comparison of reti-

nal morphology viewed by optical coherence tomography and by

light microscopy. Archives of Ophthalmology 1997; 115:

1425–1428.

35. Anger EM, Unterhuber A, Hermann B et al. Ultrahight resolu-

tion optical coherence tomography of the monkey fovea. Identi-

fication of retinal sublayers by correlation with semithin

histology sections. Experimental Eye Research 2004; 78:

1117–1125.

36. Mowat FM, Petersen-Jones SM, Williamson H et al. Topographi-

cal characterization of cone photoreceptors and the area centralis

of the canine retina. Molecular Vision 2008; 14: 2518–2527.

37. Buttery RG, Hinrichsen CFL, Weller WL et al. How thick

should a retina be? A comparative study of mammalian species

with and without intraretinal vasculature Vision Research 1991; 31:

169–187.

38. Huang Y, Cideciyan AV, Papastergiou GI et al. Relation of opti-

cal coherence tomography to microanatomy in normal and rd

chickens. Investigative Ophthalmology and Visual Science 1998; 39:

2405–2416.



use of a combined slit-lamp sd-oct to obtain anterior and posterior segment images in selected...

Documents