Optical coherence tomography (OCT) is a field in rapid evolution. Introduced in the 1990s by scientists at MIT and their clinical colleagues (led by Fujimoto and Puliafito),1 OCT has experienced explosive growth in interest along with the development of new clinical applications and instruments by manufacturers. Our interpretation of OCT images has also advanced, primarily because of improvement in the quality of the images produced, higher resolution, and accumulated clinical experience.
Like ultrasonography, OCT is a distance-measuring technology, but OCT relies on light instead of sound as the energy used for examination and light reflections instead of sound echoes as the returning signal.
The images produced by OCT are similar to those from ultrasound, but light, traveling at 1 million times the speed of sound, requires different technologies for data gathering and signal processing. OCT depends upon interferometry. As in ultrasonography, rows of OCT A-scans are grouped together to create B-scan images. The axial resolution of OCT, at 6 to 10 µm, is far greater than that of ultrasound at 100 µm. The area scanned is much smaller, usually 6 to 10 mm on a side.
Current posterior-segment OCT technology images only the posterior globe, optic nerve, and macular area. The penetration of light into tissue is limited by optical density, with the result that usually little can be seen beyond the anterior choroid.
TIME-DOMAIN, SPECTRAL-DOMAIN
Beginning in 1990, three generations of time-domain (TD) OCT have been developed and used clinically. The second and third generations each introduced significant improvements in image quality and resolution over the preceding devices. Third-generation TD OCT is currently the gold standard for posterior segment imaging. Images produced with this technology are now widely understood and interpretable by technicians and retina practitioners. Difficulties with TD devices usually involve relatively slow data acquisition, which results in patient-movement artifacts and poor image registration. Furthermore, there is no effective way on TD devices (with one exception) to compare the location of OCT images with retinal images from other diagnostic media such as fluorescein angiography.
In the past 2 years, a new OCT technology has emerged with the introduction by a number of manufacturers of spectral-domain (SD) OCT. SD OCT differs significantly from TD technology in that signals returning from the eye are scanned by a spectrometer to analyze alteration of light frequency compared with input frequency. Faster scanning and data collection are possible and result in many improvements, such as less patient-movement artifact and markedly improved image registration.
With more data acquired in each scan session, less interpolation between scans is required, making volumetric analysis and 3-D imaging possible. Measurements are reproducible within reasonable tolerances, and registration aids in establishing proper sites for measurement. Furthermore, several of the spectral-domain devices incorporate overlay software that permits correlation of OCT images with angiographic or autofluorescent photographic studies. The physician develops a better understanding of the anatomic site represented by the OCT image. Follow-up exams can be closely correlated with images from previous visits.
The Topcon 3D-OCT1000 (Paramus, NJ) is one of this new generation of spectral-domain OCT instruments. The light source is a low-coherence superluminescent diode with a wavelength in the near infrared at 840 nm and a bandwidth of 50 nm. The axial resolution is 6 µm and lateral resolution is 20 µm.
The device can produce cross-sectional B-scans, like TD OCT but with better resolution, and it can also create 3-D area scans by combining B-scans. Its scanning technology takes 20,000 A-scan measurements per second, produces a linear B-scan in less than 0.027 second, and combines them to create a 3-D area scan in less than 3.4 seconds. A 6x6-mm area scan can be created by combining 128 or up to 256 horizontal B-scans.
Point-to-point registration allows the user to pinpoint the location of retinal landmarks with comparison between OCT image and photography (Figure 1).
In addition, with 3-D image reconstruction, the 3-D area scans can be manipulated and viewed from multiple angles. Figure 2 shows views from different angles of the same 3-D OCT reconstruction of an eye with vitreomacular traction. These images were produced with the 3-D image construction feature of the 3D-OCT1000.
We used the 3D-OCT1000 in a study of 48 eyes with vitreomacular traction.2 The unprecedented visualization provided by this technology enabled us to determine the specific alterations in retinal anatomy characteristic of focal and broad vitreomacular traction.
Spectral-domain OCT also shows promise as an instrument for following and documenting the status of drusen in dry age-related macular degeneration (AMD) with volumetric analysis. Currently, a number of pharmacologic agents are being evaluated in clinical and preclinical investigations for the treatment of dry AMD. If we are to address dry AMD pharmacologically with the same degree of success we have begun to see with treatment of exudative AMD, we must have a medium capable of detecting and documenting changes in drusen. The volumetric analysis possible with SD OCT may for the first time allow us to perform automated drusen detection and quantification, although this remains to be demonstrated in clinical studies.
Some SD OCT instruments allow physicians to participate remotely in image selection. Workstations equipped with appropriate software can receive original data from the OCT instrument, permitting a physician at a remote location to monitor the images that have been taken, determining image quality and establishing whether all areas of interest have been addressed. This allows the technician or photographer taking the images to proceed with other patients without calling the physician to the instrument or printing images for review, improving work flow. Immediate 3-D imaging software also allows overall topographic map reconstruction for analysis, surgical planning, and patient education.
Optical coherence tomography, and especially SD 3-D OCT, have begun a new era in ocular imaging. Although the images are not the equivalent of pathology slides, their similarity to the appearance of retinal architecture in normal and disease states is compelling. Clinical experience and correlation with other digital imaging information over time will sharpen and improve our diagnoses as well as our understanding of many vitreoretinal diseases.
Yale L. Fisher, MD, is a Clinical Professor of Ophthalmology at the New York Hospital Cornell Medical Center, and in private practice with Vitreous-Retina-Macula Consultants of New York and a Voluntary Clinical Professor of Ophthalmology at the Bascom Palmer Eye Institute. Dr. Fisher states that he has received consulting fees from Topcon. He may be reached at phone: +1 212 861 9797; fax: +212 628 0698; E-mail: yfisher42@aol.com.
|
