Skip to main content
  • OCT: Getting the Best Images

    By Leslie Burling-Phillips, Contributing Writer

    This article is from November/December 2006 and may contain outdated material.

    Optical coherence tomography has come a long way since it was introduced in 1991 by researchers at the Massachusetts Institute of Technology. Made possible by developmental strides in the fiber optics and photonics industries, OCT first became widely available in 1996. Though a major breakthrough at the time, the emerging technology was cumbersome. “It took up to an hour to test a patient. And, often you were guessing about what you were actually looking at,” said ophthalmic photographer Patrick Saine, MEd, CRA, FOPS.

    In the last 10 years, the technology has improved, and these noninvasive, cross-sectional images have become useful for the diagnosis and monitoring of glaucoma, diabetic retinopathy, macular degeneration and macular holes, as well as other retinal diseases. The current version is now considered part of the standard of ophthalmologic care.

    “OCT has emerged as an essential imaging modality in the diagnosis and management of retinal disease,” explained retina specialist, Michael Colucciello, MD, clinical associate of ophthalmology at the University of Pennsylvania. “Being able to understand changes in the retinal architecture has allowed us to perform macular surgery better than ever and has enabled us to evaluate the results of macular surgery in a more efficient way.” 

    The Images Don’t Take Themselves

    Considering the importance of OCT in clinical decision-making, the veracity of these images is critical. This means whoever takes the images—whether an ophthalmic photographer or a technician—must know how to produce pictures of consistently high quality. “There’s a learning curve for the technology itself, and there’s a learning curve with interpreting the technology. The OCT operator obtaining the appropriate information from the patient and then the physician being able to interpret that information are two completely different learning curves that operate independently of one another,” said Mr. Saine.

    Fundus skills are a good start. Being proficient in taking fundus photographs can be helpful when learning to work with OCT imaging. It generally takes about as much training to obtain quality OCT images as it does to obtain good fundus photographs. Moreover, in addition to an intimate understanding of retinal anatomy, the photographer must have similar patient skills to perform OCT.

    “OCT imaging is very comparable to fundus photography in terms of technique and area involved. In a lot of ways the training procedures are very similar, which is why ophthalmic photographers are often the best people to train for OCT, as opposed to a technician who is accustomed to looking at the front of the eye,” Mr. Saine said. “Only half of the job gets done when you learn what all the buttons are used for and where you are supposed to be focused. The other half of the job is managing the patient so that you get a good test result. For example, having the patient look in the correct place and blink appropriately, lining the machine up properly with the eye, and knowing which of the myriad tests and settings to use.

    “There are 17 to 20 different tests currently that you can perform with OCT. Determining which test is appropriate for a particular patient is a combination of understanding the information that the physician needs and the ability to gather that information from the patient.” 

    Working With Limitations

    As with all developing technologies, the OCT does have limitations.

    Eyes must be still. “Eye movement artifacts are one of the biggest issues with OCT,” said Mr. Saine. He illustrated this point by comparing OCT with fundus photography: “Fundus photography uses a flash so all of the motion of the patient’s eye stops when the image is taken. The difficult part for the patient is that the flash is bright. On the other hand, OCT takes about a second to record the information because it uses a scanning line that crosses the retina. While it is more comfortable for the patient, the patient has to remain motionless for a longer interval of time.”

    Working with nystagmus. Although a patient may be able to remain still for the required period in most cases, challenges may arise when acquiring images from patients with nystagmus. To counter this problem Mr. Saine suggested timing the acquisition to a null point in the nystagmus. “In other words, if the nystagmus moves from side to side, you should position the acquisition to occur at the edge of the nystagmus, when the patient’s eyes are situated toward one direction or the other. Acquire the image when the eyes are most still for the longest increment of time.” He warned, however, that this could “present problems with where you are imaging with OCT. You may have to use techniques like physically turning the patient in a particular direction or other appropriate measures to aim the instrument in the right place.”

    Comparing old images to new. Another limitation to current OCT technology is that “the software doesn’t allow us to simultaneously view the acquired images from previous examinations with the current image on the screen,” said Dr. Colucciello. So, rather than being able to visualize changes in their disease progression, patients must retain a mental image when making the comparison between past and current images.

    You can’t create clarity where there is none. “One of the most important things to remember about OCT is that it’s similar to fundus imaging in that it’s optical. If you can’t see well with an ophthalmoscope or a fundus photograph, you may or may not get a good OCT image. There are times when the information you obtain from an OCT may be very faint. So it may be a ‘yes’ or ‘no’ answer that you’re looking for rather than a degree. For example, is there swelling or is there no swelling?” explained Mr. Saine. 

    The Future: Ultrahigh Resolution and Three Dimensions

    In 2001, ultrahigh resolution (UHR) OCT was first reported in the literature. This developing technology significantly improves image quality and enables definitive visualization of individual retinal layers. Clinical studies of UHR OCT are currently under way at the New England Eye Center (NEEC) in collaboration with the Medical University of Vienna.

    To date, more than 700 patients at the NEEC have been imaged using a prototype UHR OCT system. Compared with standard 10 µm resolution, UHR OCT produces improved visualization of inter-retinal layers to 3 µm. With the improved system, the ganglion cell layer, plexiform layers, nuclear layers, external limiting membrane and photoreceptor inner and outer segments can be visualized. The system also dramatically reduces eye motion artifacts, and therefore produces more accurate topographical measurements. “UHR OCT images are spectacular and really depict the retinal layers with tremendous accuracy,” Dr. Colucciello said.

    Unfortunately, current UHR OCT operates with expensive and specialized lasers and therefore is not yet practical for general use in the ophthalmology clinic. But ultimately, even three-dimensional OCT imaging will be possible.

    While OCT has become an integral part of the imaging armamentarium, it won’t become the only useful tool. “I think of OCT as an adjunct to fundus photography and fluorescein angiography, but I wouldn’t say it replaces either —it just adds information,” Mr. Saine said.


    Dr. Colucciello and Mr. Saine report no related financial interests.

    Getting the Optimal OCT

    The following tips for getting the best possible images from the Stratus OCT were compiled from suggestions provided by Michael Colucciello, MD, as well as by Marianne Whitby, senior product manager at Carl Zeiss Meditec.

    Patient Interaction

    • Clean the lens thoroughly between patients.
    • Dilation will often improve signal strength, image quality and the fundus image.
    • To improve patient fixation and reduce distraction, patch the fellow eye, particularly in patients with poor vision or when scanning the optic nerve head or retinal nerve fiber layer.
    • Instruct patients to close eyes between each acquisition of vector scans to keep the corneal surface lubricated.
    • Use artificial tears with dry eye patients.
    • Use an assistant to help with head fixation and support in cognitively impaired and physically disabled patients. Use the Fast Scan.
    • Adjust table height for wheelchair patients and perform the scan with the patient in the wheelchair.

    Improving Focus

    • When focusing the scan, the patient should be instructed where to fixate and to “keep teeth together, chin down and head still.”
    • The Focus knob should initially be set at the number that corresponds to the patient’s refractive correction, in diopters, if known. It is located on both sides of the patient module at the patient’s eye level.
    • For optimal scan quality, the Focus knob can be further adjusted to show the maximum amount of red in the OCT image.
    • Adjustment of the Focus knob will also clearly focus the dots of the scan lines. The center of the scan pattern should be aligned with the green fixation dot for macular scans. For glaucoma scans, alignthe central white Landmark light with the green dot.

    Centering Techniques

    • The Centering Line is helpful to center a scan over the fovea, optic nerve head, or other areas of interest and is particularly useful with the Fast Macula Scan and Fast Optic Nerve Head Scan.
    • In the active scanning screen, right click the Scan Image function to display the vertical line in the center of the scan.
    • Center the scan within the OCT image frame (left side of the acquisition screen). In cases of inaccurate placement, a Scan Too High or Scan Too Low message may result. In cases of gross errors (such as small pupil, media opacity or blink), a Missing Data message will appear in the box just beneath the signal strength reading. Repeat the scan in these situations.

    The Repeat Function

    • After a patient’s initial scan, use the Repeat function for subsequent visits to improve scan reproducibility.
    • Repeat Scan allows you to duplicate any saved scan group using the same set of parameters. These parameters include scan size, angle and the placement of scan, landmark and fixation.
    • When acquiring or reviewing a Repeat scan, use the Show Old Fundus Image function to toggle between the current fundus image and the saved fundus photo from the original scan. This allows for the verification of proper placement over time.

    Imaging Specific Pathologies

    • If a patient has a lesion that affects the contour of the pigment layer (subretinal pigment epithelium choroidal neovascular membrane, raised nevus, or pigment epithelial detachment, for example), it will be artificially flattened out as well in the image.
    • Print from the Scan Selection screen to obtain a large image of the unaltered cross section. The Proportional analysis option will present the scan image in its true proportions without distortion.
    • For pathologies such as macular hole, pigment epithelial detachment or epiretinal membrane, a clinician might not be as concerned with thickness mapping as with obtaining a high-resolutionimage. The standard Macula Scan, Line Scan or Cross-Hair Scan can all be used to obtain high-resolution cross-sectional images of the area of interest.
    • For some retinal conditions, such as macular holes (lamellar or full-thickness holes), AMD and conditions that may cause subretinal fluid formation, the automated analysis may fail, but the cross-sectional images are still valid and are critical for diagnosis and patient management.
    • To image macular holes, instruct the patient to look at the fixation point and “make it disappear.” The fixation point will seem to disappear to the patient when it is projected into the macular hole. Obtain the image when the fixation point disappears into the macular hole.
    • Use the Move Scan/Landmark function to image the fovea in patients with fixation problems and central scotoma (such as AMD patients).
    • For patients with central scotoma, instruct patient to look “in the center” of the circular alignment scan using the Standard (rather than Fast) mode.
    • Use Fast Scan to obtain an image in patients with media opacity (e.g., corneal edema, dense cataract or vitreous hemorrhage).
    • In patients with myopia greater than –12 D, or posterior staphyloma, use the Refraction Correction Wheel to obtain “as much minus as possible.”

    Memory Stick

    • A USB drive is a convenient way to transfer data or capture images for presentations.
    • A USB hub can be attached to the printer cable without removing the back cover of the Stratus computer.