A Low-Cost and Simple Imaging Technique of the Anterior and Posterior Segments: Eye Fundus, Ciliary Bodies, Iridocorneal Angle

Abstract

purpose. The authors recently used topical endoscopy to image the mouse eye fundus. Here, they widened the field of application for this ophthalmologic tool, imaging both the posterior and the anterior eye segments in larger animals commonly encountered in research laboratories and veterinary clinics.

methods. Pupils were dilated, and local anesthetic and gel were applied to the animal cornea. The endoscopic probe was placed in contact with the cornea of conscious rats, sedated cats and dogs, anesthetized sheep, and nonhuman primates.

results. High-resolution digital images of the eye fundus were obtained in all investigated animals using the endoscopic probe along the eye axis. Arteriovenous filling time was monitored with fluorescein angiography in pigmented rats. The retinal periphery and ciliary bodies could be visualized with the probe placed at an oblique angle. The probe was inclined further to observe the iridocorneal angle such that the pectinate ligaments could be seen at high resolution in cats. The authors used the probe on eyes with retinal detachment, luxation of a cataractous lens, and pigment infiltration in the iridocorneal angle, demonstrating its potential use in eye diseases.

conclusions. This topical endoscopic technique provides a unique tool for single eye examinations. The authors obtained a circular view of the anterior (iridocorneal angle) and the posterior (fundus) eye segments from all animal species studied. This technique is inexpensive and easy to use. It can be easily moved to the eye of the patient who cannot move to stand in front of classic apparatus, offering new opportunities in ophthalmology.

Imaging eye structures has become essential in diagnostic examinations and follow-up of retinal diseases or lesions in humans and animals. Imaging the eye fundus, for instance, detects retinal detachment, retinal atrophy, and blood vascular occlusion, and examination of the iridocorneal angle can reveal factors underlying an increase in intraocular pressure leading to glaucoma and ganglion cell degeneration. 

Eye examinations of the cornea, anterior chamber, lens and posterior chamber, and vitreous are usually carried out with a slit lamp apparatus. Images of the eye fundus can then be obtained using indirect ophthalmoscopy or confocal scanning laser ophthalmoscopy (cSLO). ¹ These techniques not only provide reflection images of the retina, they generate fluorescent images from natural pigments (autofluorescence) or fluorescent dyes (e.g., fluorescein and indocyanine green) administered to the patient or animal. Such dye-induced images are particularly useful for assessing blood circulation and blood vessel permeability in the retina and the choroid. Anatomic sections of the in vivo retina can be obtained by optic coherence tomography (OCT), a technique recently associated with cSLO. ² These techniques all use eye optics to visualize the fundus without any contact with the cornea. However, they do not enable the far retinal periphery to be visualized. OCT can provide images of the iridocorneal angle of the eye and ciliary bodies. ^{3 4} Ultrasound biomacroscopy (UBM) can also provide images of the iridocorneal angle and of the ciliary bodies in humans ⁵ and mice. ⁶ To visualize these eye structures for surgery, a goniolens must be used on the cornea to obtain overall circular visualization through indirect ophthalmoscopy. ^{7 8} However, this approach does not allow a particular area of interest to be targeted by the clinician. 

We recently described a new endoscopic technique to obtain eye fundus images in the mouse retina. ⁹ In this study, we extended this low-cost technique to use in larger mammals, including nonhuman primates. Furthermore, we used it to visualize other eye structures, including the iridocorneal angle, in a single examination by simply changing the angle of the endoscopic probe on the cornea. 

Materials and Methods

Animals were handled in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and with French laws and regulations. Three dogs (one healthy, two with abnormalities) and two cats were examined at the Eye Veterinary Clinic (SR) after consent was obtained from the owners. Only one monkey, one marmoset, and one sheep were used for imaging because images could be obtained at a first glance without any difficulties in these adult animals (these species were selected because they have been previously used in research and regulatory studies or because they are often encountered in veterinary eye clinics). 

The endoscopic imaging system has been described in detail elsewhere. ⁹ Briefly, the technique is based on an endoscope with parallel illumination and observation channels connected to a digital photographic camera or video camera. We used an endoscopic probe (2.7-mm diameter, 10-cm length, 1218AA; Optique Hopkins; Karl Storz, Tuttlingen, Germany) for rats, cats, dogs, and monkeys. For screening procedures, the endoscope was clipped to a 640 × 480 dedicated video camera (Starcam; Karl Storz; Fig. 1 ) connected to a monitor. Images were also obtained by connecting the endoscope (AIDA compact II; Storz) through an adapter to either a reflex digital camera with a 6.1 million pixel charge-coupled device (CCD; D50 with Nikkor AF 85/F1.8 D objective; both from Nikon, Tokyo, Japan) or to a video camera (GV-D800; Sony, Tokyo, Japan). The light source was a xenon lamp (reference 201315-20; Storz) linked to the endoscope by a flexible optic fiber. For fluorescein angiography, a blue filter (20100032; Storz) was inserted in front of the optic fiber, and a yellow filter (20100033; Storzsd) was inserted in front of the camera. 

Adult rats were provided by Janvier (Saint-Ile le Genest, France) or Charles Rivers (L’Arbresle, France). Monkeys were supplied by Harlan (Loughborough, Leicestershire, UK) and sheep by INRA (Jouy en Josas, France). All experiments were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. All animals received veterinary care in compliance with the Guide for the Care and Use of Laboratory Animals, and experimental procedures were approved by the institutional ethics committee. 

Rats were anesthetized by intraperitoneal injection (0.8–1.2 mL/kg) of a solution containing ketamine (40 mg/mL) and xylazine (4 mg/mL Rompum) so that retinal detachment could be performed. Pupils were dilated using tropicamide (0.5%; Novartis Pharma SAS, Créteil, France). A microinjector (UltraMicroPump II-Micro 4; World Precision Instruments, Sarasota, FL) was connected to a syringe filled with 12 mg/mL sodium hyaluronate (Amvisc; Bausch & Lomb, Rochester, NY). Under a stereomicroscope (Carl Zeiss, Oberkochen, Germany), a 30-gauge needle (Becton Dickinson SAS, Rungin, France) was inserted approximately 2 mm posterior to the limbus, through the sclera, and directed toward the posterior pole of the eye, into the subretinal space. Care was taken not to damage the lens during needle penetration. Sodium hyaluronate was then slowly injected under visual control into the subretinal space with the injector, thus detaching the retina from the underlying RPE. During detachment a hole was created within the detached retina to allow long-lasting detachment. Retinal detachments were created only in the left eye of each animal on approximately half the retina; the right eye served as the control. 

Retinal vein hemorrhages were created on dilated eyes of anesthetized rats by laser photocoagulation. ¹⁰ The fundus was visualized through a lens (SuperPupil; Volk Optical, Mentor, OH) with a slit lamp coupled to a laser (wavelength 534 nm; crystal focus; Alcon, Forth Worth, TX) used to rupture major veins (spot size, 50 μm; duration, 0.5 seconds; power, 100–150 mW). 

Cats and dogs were conscious but sedated with medetomidine chlorohydrate, (Domitor; Pfizer, Paris, France) at a dose of 0.17 mg/kg. Other animals were anesthetized as follows: the marmoset with a combination of ketamine chlorohydrate (Imalgene 500; Merial, Lyon, France; 20 mg/kg intramuscularly) and medetomidine chlorohydrate (Domitor; Pfizer; 0.1 mg/kg intramuscularly); the monkey (Macaque Cynomolgus or Macaca fascicularis) with ketamine (10 mg/kg intramuscularly [Imalgène 1000; Merial] + 0.2 mg/kg midazolam intramuscularly [Hypnovel; Roche Neuilly sur Seine, France]), and the sheep with sodium thiopental (10 mg/kg intravenously [Nesdonal; Merial]). The monkey and sheep were endotracheally intubated and placed on mechanical ventilation (2% isoflurane in 100% O₂). Animals were monitored by electrocardiography and by measurements of invasive blood pressure, end tidal CO₂, and body temperature. 

Pupils were maximally dilated with topical 1% tropicamide (Mydriaticum; Théa, Paris, France). This was complemented by phenylephrine hydrochloride ophthalmic solution (Neo-Synephrine 10%; Bayer, Pittsburgh, PA) in monkeys because the maximal concentration of tropicamide solution, limited to 2% in Europe, does not achieve complete pupil dilatation in these animals. In conscious animals, topical oxybropucaine was used for corneal anesthesia (all eyedrops were from Novartis Ophthalmics, Rueil Malmaison, France). We applied a corneal gel (Ocrygel; Laboratoire TVM, France) to protect the corneal epithelium and to create an interface that would improve the quality of the image. For fluorescein angiography in rats, 50 μL of 10% sodium fluorescein solution (Ciba Vision, Duluth, GA) was injected intraperitoneally, and the fundus was examined with appropriate excitation and barrier filters. 

Two operators were needed to examine conscious rats, and only one was required to study sedated or anesthetized animals (monkeys, cats, dogs, and sheep). Focus and illumination were adjusted by direct examination of the eye structures through the camera. Digital images and video were captured on the camera or Sony video camera. Images transferred to a computer were processed and presented with the use of Adobe Photoshop (Adobe Corporation, Mountain View, CA). 

Results

We could observe the fundus with the endoscopic probe placed on pigmented (Fig. 2A)and albino (Fig. 2B)rat eyes. Blood vessels were visible in both strains. For fundus imaging in albino rats, we had to decrease the amount of incident light and the duration of exposure to avoid overexposure. In pigmented animals, ganglion cell axon bundles radiating from the optic nerve disc were detected as white fibers running on top of the retina (Fig. 2A , arrow). The retinal pigment epithelium was also visible below the retina. The technique is now used for weekly examinations of eye fundi in a retinal prosthesis project instead of the less robust scanning laser ophthalmoscope. ¹¹ Many users have been able to produce images without extensive training, though specific training is required for image interpretation. We were able to focus on the eye periphery by orienting the endoscope at the surface of the eye at large angles relative to the central optic axis of the eye. Examples of the peripheral retina with the ciliary body visualized in an albino and a pigmented rat are shown in Figure 3 . This technique is now used in a number of projects requiring imaging of the anterior and posterior segments in rats; we did not find any conditions preventing its use other than opacities of the cornea or lens. 

We obtained images of the eye fundus for all species studied, including sheep, cat, dog, and marmoset (a small nonhuman primate; Figs. 2C -F). In the sheep, cat, and dog, the eye fundus is composed of two different light-reflecting areas, the tapetum lucidum (TL) and the tapetum nigrum (TN). ¹² The dense pigmentation of the tapetum nigrum prevented clear visualization of blood vessels and of the surrounding retinal tissue in this area (Figs. 2C 2D 2E) . The black appearance of the tapetum nigrum produces a seemingly halved image of the eye fundus. In all animals, the optic nerve head and blood vessels were easily visible. Images in the marmoset clearly detected the fovea (Fig. 2F , arrow). The thinning of the retinal pigment epithelium was visible in an old cat, viewed as brighter or hyperreflective areas of the eye fundus (Fig. 2D , arrow). We used the same apparatus as that used in mice to examine ciliary bodies in larger animals, such as monkeys (Fig. 3C) , by applying an indentation on the eye. Thus, our endoscopic technique, first reported in mice, can also be applied to larger animals to visualize their eye fundus and age-related changes. 

Fluorescein was administered to rats; we carried out fluorescein angiography on conscious animals. Figure 4illustrates different stages of arteriovenous filling in the rat. In the first examination (Fig. 4A) , only blood arteries were fluorescent, whereas, at later stages, retinal arteries and veins were visible (Figs. 4B 4C) . Only the venous wall appeared fluorescent during precocious arteriovenous filling (Fig. 4B) , whereas veins were homogenously fluorescent at later stages (Fig. 4C) . The technique also allows areas of particular interest, such as the ciliary bodies (Figs. 4D 4E)or the periphery (Fig. 4F) , to be targeted. This technique seems to be adequate for the study of blood vessel occlusion and permeability. 

In addition to imaging the eye fundus, we used the endoscope to visualize structures of the anterior segment. Figure 5illustrates visualization of the anterior chamber in several species, including cat (Fig. 5A) , dog (Fig. 5B) , rat (Fig. 5C) , sheep (Fig. 5D) , and monkey (Fig. 5E) . The iridocorneal angle could be examined with great resolution using the endoscopic probe oriented on the cornea to obtain a wide angle to the optic axis of the eye; this showed, for example, the pectinate ligaments of the trabeculum at high resolution in cats and dogs (Figs. 5A 5B) . Indentation was needed in monkeys. This technique could thus be used by a clinician to observe eye structures at the front and the back of the eye during a single eye examination. 

We examined animals with ophthalmologic lesions to evaluate the potential use of our endoscopic technique in routine practice. In rats subjected to experimental retinal detachment, the detached area was clearly visible (Fig. 6A) . Furthermore, we were able to detect areas with incisions into the retinal tissue through observation of the detached areas at high resolution (Fig. 6B) . Hemorrhages were also clearly detected after retinal detachment or vascular occlusions in rat eyes (Fig. 6C) . Among the larger animals, a dog affected with cataractous lens luxation was examined in the eye veterinary clinic (Figs. 6D 6E) . We observed the opaque lens at the front of the eye, with elongated ciliary processes (Fig. 6D) . We oriented the endoscope to obtain clear images of the eye fundus in this animal despite the deteriorated optic axis (Fig. 6E) . Another dog had increased intraocular pressure and was examined in the eye veterinary clinic. The iridocorneal angle exhibited substantial pigment infiltration and a diminished trabeculum meshwork (Fig. 6F) . These examinations demonstrated that this technique can be used to investigate disorders of the eye. 

Discussion

We previously demonstrated the use of the endoscopic technique on mouse retina, focusing on the posterior segment and, in particular, comparing the use of different wavelengths for examination of blood vessels. ⁹ In this study, we demonstrated its potential use on larger animals, including nonhuman primates such as the marmoset, used in regulatory toxicologic studies. We have illustrated that the technique can be applied to different eyes with various shapes and optical properties. Furthermore, we found that the same tool can be used to visualize not only the posterior segment but also the iridocorneal angle in the same examination and in a reproducible manner. These findings, therefore, confirm that this technique may be beneficial in vision research, toxicologic and pharmaceutical studies, and veterinary ophthalmic examinations. They also raise the question of whether the technique may be used for clinical application in humans. However, this would require that the light source be adapted by filtering UV light and that the light intensity be limited during the eye examination by introducing a flash of light for image acquisition. These specifications are not required in an otoscope but would be necessary for an “ophthascope” to limit potential light damage of retinal cells and patient discomfort. 

Indirect ophthalmoscopy, another technique used for imaging the eye, allows visualization of the fundus without details of the retina. The technique requires an extensive training period for successful visualization of the retina and does not produce high-quality images. Biomicroscopy with slit-lamp examination also requires a long training period to visualize the anterior chamber, the lens, and the vitreous; this technique is difficult to use in small species. ¹³ Although new hand-held slit lamps have reduced significantly the cost of the equipment, they cannot provide the clear, recorded images obtained with our endoscopic techniques. Furthermore, such techniques requiring relatively cheap equipment do not generate fluorescent images for fluorescein angiography. Our endoscopic technique has the capacity to generate images for fluorescein angiography; thus, it could be adapted to image fluorescence generated by natural pigments such as lipofuscin or by other injected dyes. Previous studies requiring well-focused and detailed images of the animal eye fundus had to make use of a retinograph, an angiograph, or a cSLO, calling for expensive equipment. ^{14 15} SLO images can be produced from different layers of the retina, providing detailed information that cannot be generated by our endoscopic technique. However, the cSLO is a large device that requires tight alignment of the laser beam, and it cannot be easily moved. Even with such sophisticated imaging apparatuses, it is difficult to obtain images from peripheral areas and ciliary bodies. By contrast, our endoscopic technique incorporates all directions in a hemisphere by simply positioning the endoscopic probe in the desired direction. For instance, a previous investigation of lens accommodation in monkeys involved the removal of the iris so that the movement of ciliary bodies could be studied in vivo. ^{16 17} Changes in the iris and ciliary bodies occurring in dystrophic cats have previously only been reported after histologic observations. ¹⁸ OCT and UBM also provide in vivo images of the iridocorneal angle and ciliary bodies. ⁴ Our endoscopic technique offers direct visualization of the various eye structures with a tool that is easy to handle. Training is needed only to reproducibly visualize the same point under the same axis and to improve the quality of acquired images for their online interpretation. 

Rats, cats, and dogs investigated in this study had smaller axial eye lengths than humans, whereas those of sheep were larger (human, 23.5 mm; rat, 6.3 mm; cat, 21.3 mm; dog, 21.73 mm; sheep, 26.85 mm). Therefore, the technique could also be suitable for human patients in ophthalmology clinics. Such an application could be beneficial to examine the anterior segment, for instance. Examination of the iridocorneal angle is usually performed with a goniolens, which enables the investigator to have a circular view of the iridocorneal angle in an animal. ¹⁹ Such a technique, however, does not allow specific areas to be widened for a detailed analysis; rather, images must be produced by indirect ophthalmoscopy requiring extensive training (e.g., for localization of partial closure of the iridocorneal angle). With our endoscopic technique, we observed the iridocorneal angle with a wider view of the pectinate ligament at the trabeculum. These observations could facilitate the diagnosis of glaucoma, especially acute glaucoma with a closed angle, though it may be more difficult to obtain quantified measurements with this technique than with OCT and UBM. It could also be used to examine anterior iridocorneal synechia, a risk factor for chronic glaucoma. ⁷ Coupling an argon laser to the endoscope may provide an alternative method to achieve laser treatments of the iridocorneal angle to the current method of coupling the argon laser to the slit lamp and using the goniolens. The goniolens approach may, however, remain superior for stabilizing the eye image for the precise location of laser spots. Our endoscopic technique thereby offers a new approach to visualize the iridocorneal angle in great detail and potentially to treat trabeculum disorders by orienting the endoscopic probe. A laser-coupled endoscope could also be adapted for retinal treatments. Furthermore, other lasers could be coupled to the endoscope, such as the YAG laser for the treatment of postoperative cataracts. 

Previously, most imaging techniques of the eye fundus have not required mechanical contact with the cornea; our technique requires contact with the tip of the endoscopic probe. This requirement may hinder the use of this technique in ophthalmology and could alter image acquisition by excessive pressure on the cornea. Examination of the iridocorneal angle with the goniolens also requires corneal contact. The contact could cause abrasions or irritations of the corneal epithelium despite the presence of the gel on the cornea. The risk of damaging the corneal surface can be prevented with the use of a corneal lens, previously used for examination of the eye fundus in the marmoset (Tessier Y, personal communication, May 30, 2007). To limit potential lesions by the endoscopic probe, a round tip could be designed, thereby removing edges. Alternatively, a corneal lens could be used to provide fixed positions for the endoscopic probe, or a removable tip could be placed around the probe. These solutions would also provide sterile conditions without the need to sterilize the entire endoscopic probe. 

Imaging techniques that do not require corneal contact may cause dehydration of the cornea during sedation or anesthesia if eye drops are not continuously applied to the cornea to replace the tear film. Such an optical interface often reflects the incident light, thus generating a dot artifact on the acquired image. Neutralizing this interface with the gel eliminates this artifact and improves the general optic quality of the system, generating an improved image of the eye structures. This endoscopic technique could be combined with another imaging technique, such as the cSLO, OCT, or adaptive optics imaging, to further increase their already high resolution or to extend application to examination of the iridocorneal angle. 

As mentioned, a complete eye examination of the fundus and iridocorneal angle can be performed with our endoscopic technique by simply changing the orientation angle of the endoscope on the corneal surface. Our endoscopic technique is easier to learn than other available techniques for imaging the eye fundus. Because of the absence of focal distance between the eye and the front lens and, thus, the absence of image inversion, only a few hours are required to learn to use the technique independently. The apparatus is small and easy to use and may be particularly beneficial for imaging eye structures during surgery, such as for intraocular hemorrhages. Materials providing good-quality images of the eye fundus are generally heavy, relatively expensive, and difficult to move. This is particularly true for the SLO, which requires laser alignment after each movement. By contrast, all the pieces of equipment required for our endoscopic technique fit in a case (Fig. 1)and can operate on a battery. Therefore, this inexpensive and portable device may be adapted not only for research laboratories and ophthalmology clinics but also for the performance of full ophthalmologic examinations in remote areas. Further improvements are needed, however, to reduce the light intensity for examination by coupling the apparatus to a highly sensitive camera. 

 Supported by the Fondation Ophtalmologique A. de Rothschild (Paris), INSERM, the University Pierre and Marie Curie (Paris 6), Assistance Publique-Hopitaux de Paris (AP-HP), Fédération des Aveugles de France, and European Economic Community Grant EVI-GENORET-512036. The endoscope was paid for by the Association Sportive de Vialas and la Caisse d’Assurance Maladie des Professions libérales.

 Submitted for publication October 18, 2007; revised April 3 and May 14, 2008; accepted September 3, 2008.

 Disclosure: J.-L. Guyomard, (P); S.G. Rosolen, (P); M. Paques, (P); M.-N. Delyfer, (P); M. Simonutti, (P); Y. Tessier, (P); J.A. Sahel, (P); J.-F. Legargasson, (P); S. Picaud, (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.

 

Presentation of the experimental setting for the eye examination on a cat. The investigator is holding the endoscopic probe connected to a light source and to a video camera with a screen display showing the eye fundus. The position of the endoscopic probe on the cornea is enlarged in the inset. When the orientation of the endoscopic probe is changed at the surface of the cornea, the changes in the video images can be followed online.

Figure 2.

 

Eye fundus images obtained with the endoscopic technique. (A, B) Eye fundus examination in a pigmented (A) and in an albino (B) rat. Note in the pigmented animal the white ganglion cell axon bundles radiating from the optic nerve (A, arrow). (C–D) Eye fundi in large animals: sheep (C), cat (D), dog (E), and marmoset (F). In the cat, dog, and sheep, eye fundi are separated in two areas: tapetum lucidum (TL) and tapetum nigrum (TN). On the cat eye fundus, hyperreflective areas result from thinning of the retinal pigment epithelium (D, arrow). The fovea is clearly seen on the marmoset eye fundus (F, arrow).

Figure 3.

 

Images of the ciliary bodies in an albino rat (A), a pigmented rat (B), a marmoset (C), and a monkey (D). The ciliary body is indicated by an arrow in the marmoset.

Figure 4.

 

Images of fluorescein angiography obtained with the endoscopic technique in rats. (A–C) Sequence of arteriovenous filling times. Immediately after fluorescein injection, only arteries had become fluorescent (A). Subsequently, precocious filling of the veins was indicated by the fluorescence restricted to the vein walls, (B) whereas later, arteries and veins are homogeneously fluorescent. (D–F) Fluorescein angiography of the ciliary bodies (arrow) and peripheral retina. Ciliary bodies (D, E) and small blood vessels (F) are clearly fluorescent at the retinal periphery.

Figure 5.

 

Images of the iridocorneal angle with the endoscopic technique in cat (A), dog (B), rat (C), sheep (D), and monkey (E). In the cat and dog, the pectinate ligament (PL) can be clearly seen in the iridocorneal angle. Images also display the iris and pupil.

Figure 6.

 

Images of eye structures with the endoscopic technique in abnormal eyes. (A, B) Eye fundi showing retinal detachment in rats. (A) The detached retina is hyperreflective and less well focused, and its limits are underlined by two arrows. (B, arrow) The detached retina can be examined to localize holes. (C) Visualization of laser-induced hemorrhages in the albino rat. (D, E) Images of eye structures in a dog with a luxation of a cataractous lens (CL). (D) In the anterior chamber, elongated ciliary processes (CP) can be seen as well as the cataractous lens. In the posterior chamber (E), the retina can still be nicely visualized by selecting an appropriate angle for the endoscopic probe. (F) Intraocular hyperpressure syndrome showing pigment infiltration in the iridocorneal angle.

The authors thank the staff of the Institut Mutualiste Montsouris Recherche (Paris, France) for technical help.