7 Anterior-Segment Exploration with Optical Coherence Tomography
Until recently, most efforts in ocular imaging concerned posterior-segment exploration. One of the most important events was the commercialization of the Carl Zeiss Meditec Stratus optical coherence tomography (OCT 3–820 nm wavelength), which made it possible to visualize different layers of the retina with a great deal of precision (resolution 3–4 µm). Efforts are under way to develop increasingly precise scanning of the neurosensorial layers using three-dimensional visualization (Carmen Puliafito Innovators Lecture, American Society of Cataract and Refractive Surgery, Washington, 2005).
Because it is possible to observe the anterior segment directly using the slit lamp, extensive research has not been a priority. In daily practice, ultrasonic evaluations of corneal pachymetry and anterior-chamber depth along the optical axis of the eye have been considered sufficient. With the development of sophisticated surgical techniques, however, it became essential to obtain elaborate static and dynamic measurements of the anterior segment to meet modern safety requirements. Here one can briefly refer to X-ray imaging of the eye (conventional X-ray, MRI (Magnetic Resonance Imaging), TDM (computer tomography)) because its use in daily practice is unlikely to undergo further development. The choice now lies between optical and ultrasonic exploration of the anterior segment. Slit-lamp images are simply frontal images with a subjective estimation of a few external measurements of the eye. Development of the Scheimpflug technique with oblique images resulted in a new capability to evaluate the distances in the eye’s anterior segment along different optical sections. The major drawbacks of this technology are a difficult mathematical reconstruction and scleral overexposure when taking photographs. In particular, the whole of the angle area is masked by this overexposure, and the fine structures are indiscernible (i.e., the scleral spur, iridocorneal sinus).
The idea of using infrared wavelengths in optical coherence is expanding rapidly (IOLMaster, Visant OCT; Carl Zeiss Meditec). 1 , 2 About 10 years ago, Izatt and colleagues 3 suggested using the OCT for anterior-segment imaging. Reflection of the infrared light rays is captured and analyzed by an optical sensor, and appropriate software readjusts the dimensions of the images by erasing distortion errors attributable to different corneal optical transmission differences. Measuring software capable of evaluating the distance between two points, the curvature radius and angles, is also integrated.
Ultrasonic exploration of the anterior segment appears to have reached its limits, whether in ultrasound biomicroscopy or ultrahigh-frequency ultrasound equipment (Artemis; Ultralink, LLC). Today, resolution is identical to 1310-nm wavelength anterior segment OCT, which is available on the market (15–20 µm for axial resolution, 50–100 µm for transverse resolution). Manipulation is fairly complex, and even if some ultrasonic measurements are used as references to calibrate a certain number of instruments, there is no certainty of exact in vivo or ultrasonic measurements. However, the error can be considered relative as long as the reference scale remains constant with each device and technology.
7.1 Anterior-Segment Exploration
7.1.1 Time-Domain Anterior-Segment OCT
The anterior-segment time-domain OCT (Visante, Carl Zeiss Meditec) is on the verge of being commercialized, and we have been fortunate to be able to use it for 2 years as a prototype. 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 The equipment uses a 1310-nm wavelength, but in its present form, the infrared light is blocked by pigments; however, the nonpigmented opaque structures are permeable, and images can be obtained through a cloudy or white cornea, through the conjunctiva, and through the sclera. Axial resolution is 18 µm and transverse resolution 50 µm. The procedure is noncontact and quite easy; therefore, a technician can be rapidly trained to carry out the examinations. It is possible to choose the axis to be explored or carry out an automatic 360-degree exploration along the four meridians.
An optical target can be focused or defocused with positive or negative lenses. Natural accommodation can be stimulated, and anterior-segment modifications during accommodation can be explored in vivo. Image reconstruction software has been criticized, but in our experience, we have been able to show that the sections obtained were reproducible. We believe this notion of reproducibility is quite important. A few errors regarding the precision of the readings may be important, but as long as the reference scale remains constant and the areas explored can easily be found during successive examinations, these errors can be considered relative (Fig. 7.1 , Fig. 7.2).
7.1.2 Static Measurement of the Anterior Segment (Fig. 7.3)
With the Visante OCT prototype at our disposal, we were able to explore hundreds of eyes and to evaluate the different measurements of the anterior segment. We were able to show a new notion: in most cases, the anterior chamber was not a circle. We were able to prove in vivo that in 75% of cases, the internal vertical diameter of the anterior chamber was larger than the internal horizontal diameter by at least 100 µm. Using the Artemis instrument (ultra high-frequency ultrasound) on cadaver eyes, this notion had already been put forward by Liliana Werner. This discovery has essential implications when an anterior chamber phakic or pseudophakic angle-supported implant is scheduled. The measurement obtained with the Visante OCT is much more precise than the white-to-white evaluation that was previously used, and it is more precise and much easier to acquire than the anterior-segment images obtained using the classic B-scan instruments, even the most recent ones. Furthermore, with ultrasonic scanning devices, the water bath placed before the patient’s cornea makes the examination difficult, and because there is no fixation point, it is not possible to visualize the optical axis. The optical axis, which is a fixed reference point, allows one to know with certainty the position of the examined optical section. Thus, with the Artemis, there is lack of reproducibility.
7.1.3 Dynamic Evaluation of the Anterior Segment
Fig. 7.4 comprises images of the eye of a 10-year-old child with 10 diopters of accommodation; the images speak for themselves. Distortion of the anterior surface of the crystalline lens, myosis, and modifications to the anterior chamber depth during accommodation can be observed. This shows that in young subjects the anterior segment of the eye is quite dynamic; for 1-diopter accommodation, there is approximately a 30-µm forward thrust of the crystalline lens’ anterior pole. Therefore, it is usual to observe a 100- to 200-µm variation of the anterior chamber in a young subject (a candidate for a phakic implant).
With aging, crystalline lens flexibility reduces and fewer modifications to the anterior segment occur during accommodation; however, the anterior chamber becomes flatter as the crystalline lens’ anterior pole moves forward by about 20 µm per year. Specific software could perhaps be used to simulate the aging of the anterior segment and thus help to explain that as the crystalline lens thickens with time, contact between the crystalline lens and all models of phakic implants is probable, regardless of their type of fixation in the anterior segment.
7.1.4 Evaluation of the Crystalline Lens with the Anterior Chamber OCT
Fig. 7.5a shows that by focusing behind the iris, it is possible to observe the entire thickness of the crystalline lens.
Fig. 7.5b shows the image of a 2-year-old child with Peter syndrome and lens-cornea adherence; the different layers of the crystalline lens are perfectly visible.
With current equipment, densitometric evaluation of the crystalline lens is not possible, but it is hoped that with ongoing studies, such analysis will be possible shortly.
7.1.5 Pseudophakic Artificial Crystalline Lens
Fig. 7.6 shows two examples of a piggyback implant, and Fig. 7.6a shows perfectly well-adapted piggyback intraocular lens with the main lens behind and the secondary lens in front. The interface between the two lenses is virtual but perfectly visible on the images. In this case, there was no proliferation of unwanted deposits on the interface. In Fig. 7.6b, perforating keratoplasty was carried out on a corneal pseudophakic edema. Using the slit-lamp, an after-cataract was diagnosed postoperatively. In fact, after an examination with the OCT, we noticed a proliferation of newly formed tissues between the two piggyback implants. We changed our diagnosis of after-cataract to intralenticular proliferation.