Improving Cornea Care With Fourier Domain OCT

Learn how to use the RTVue-100 Cornea-Anterior Module for LASIK, keratoconus, glaucoma and other conditions.

By David Huang, M.D., Ph.D.

One of the most recent advances in diagnostic imaging is Fourier domain optical coherence tomography (OCT), a system that offers dramatic improvements in speed and resolution, and moves us closer to the original goal of using OCT to evaluate pathology at the microscopic level without relying on biopsies.

I'm particularly excited about the RTVue-100 (Optovue Inc., Fremont, Calif.), the first Fourierdomain OCT system to be introduced in the United States. It's the only commercially available OCT system that's FDA-approved for both corneal and retinal imaging. Here, I'll describe the advantages of the RTVue-100 and the benefits of the system's corneaanterior module (CAM), which received FDA clearance in September 2007.

Dramatic Advances in OCT

My work in developing OCT dates back to 1991, when my colleagues and I described the technology in the journal Science.¹ Since then, the performance of OCT has improved greatly. The most important advance has been the development of very high speed Fourier domain OCT technology.

I can best describe the benefits of Fourier domain by comparing it to a time domain approach. Consider the following:
■ Speed. The Fourier domain RTVue system provides 26,000 axial scans per second, compared to best commercial time domain anterior segment OCT, which provides 2,000 axial scans per second. This 13-fold speed improvement is comparable to going from a propeller biplane to a jet airplane. Fourier domain OCT achieves this higher speed by simultaneously collecting an entire 2,000-pixel axial scan. In contrast, time domain OCT acquires the axial scan 1 pixel at a time. In the time this older technology takes to obtain one axial scan, we can acquire a complete image using Fourier domain OCT.
■ High-definition Images. Because of the higher speed, the RTVue system provides a high-definition (1,000 line) image in 0.04 seconds, freezing out most biological motion errors. This is far superior to the older time domain technology, which typically obtains a medium-resolution (500 line) image in a quarter of a second.
■ Resolution. The 5-micron axial resolution produced by the RTVue-100 is more than three times higher than that of time domain OCT.
■ Versatility. The RTVue can be used for both retinal and corneal imaging. Corneal scanning is achieved with the use of adapter lenses. One adapter lens is for high magnification — or high transverse resolution — imaging. The other adapter lens is for wider field imaging.

Advantages of Corneal Scanning

The cornea-anterior module of the RTVue-100 can assist with the following tasks:
■ Pre-LASIK screening of patients to rule out keratoconus
■ Postoperative evaluation of LASIK flaps and stromal bed
■ Corneal power measurement
■ Treatment of corneal scars
■ Evaluation of tear film and meniscus
■ Evaluation of keratoconus prior to implantation of Intacs intracorneal implants.

As seen in Figures 1 and 2, the image produced by the RTVue-100 is an average of 16 frames. The higher speed of the technology suppresses speckle noise and makes fine features more prominent, allowing us to measure the tear film. We also can measure the lower tear meniscus, another indicator of dry eye.

Figure 1. The RTVue-100 scan can differentiate a contact lens, Bowman's layer and Descemet's membrane.

Figure 2. The RTVue-100 provides precise images (left) and axial scans (right). The peaks in the scans indicate the air/tear and tear/contact lens interfaces before and after artificial tears are instilled. Because both types of peaks are combined by the baseline scan, an indirect measurement is used to calculate the tear film thickness. The tear and epithelium thickness from the baseline scan are measured. Then the epithelium thickness is subtracted to calculate tear film thickness.

Post-LASIK Applications

The higher resolution of the RTVue-100 also allows us to more precisely visualize the flap after LASIK. Visualization is more challenging with the use of a single frame, especially several months after surgery. By averaging frames, we can see the very fine interface.

Keratoconus and Intacs

We're accustomed to using corneal topography and slit-scanning technology to monitor keratoconus patients. Slit-scanning technology, such as the Orbscan (Bausch & Lomb, Rochester, NY), provides a pachymetry map. But the Orbscan tends to underestimate corneal thickness in the setting of even mild corneal haze and scar, including keratoconus. Errors in the Orbscan pachymetry map can be significant when determining if a keratoconus patient is a candidate for corneal implants.

Figure 3. The minimum corneal thickness determined by Orbscan II for a keratoconus patient is 339 microns, ruling out the use of Intacs because 400 microns is the threshold for the procedure.

A comparison of Orbscan II and RTVue-100 scans in Figures 3 and 4 demonstrates the importance of this issue. The more precise measurement of minimum corneal thickness provided by RTVue-100 is consistent with what's found during a slit lamp examination, qualifying the patient for treatment with Intacs intracorneal implants (Addition Technology Inc., Des Plaines, Ill.). The treatment would have been ruled out for the same patient if we'd relied solely on Orbscan II, which showed a minimum thickness below 400 microns.

The pachymetry map produced by RTVue-100 also provides much more information for diagnosing keratoconus and forme fruste keratoconus. It's important to rule out these conditions in the preoperative evaluation of prospective LASIK patients. Several diagnostic parameters capture the characteristics of keratoconus and offer good diagnostic accuracy. To detect the asymmetric thinning that's typical of keratoconus, look at the difference between the inferotemporal and superonasal octants, and the difference between the inferior and superior octants. Keratoconus eyes typically have a thinner minimum thickness. The minimum minus median thickness is an even more specific parameter that reveals focal thinning rather than general thinning.

Figure 4. The measurement using RTVue-100 is 403 microns, qualifying the same patient for Intacs.

Precision of the RTVue-100

The precision of the RTVue-100 is achieved because of its speed. When using the RTVue-100, we're sampling more of the eye in a very short period. This translates to less motion error. With time domain technology, at 2,000 scans per second, the corneal power measurement had a reproducibility of about 0.8D (pooled standard deviation). That's not as good as corneal topography placed on placido rings, which has a reproducibility of 0.2D to 0.3D.

At 26,000 scans per second, the RTVue-100 can measure corneal power as precisely as placido rings. But unlike Placido ring-topography, which measures the anterior corneal surface only, the RTVue can measure the front and back surfaces of the cornea. This has great implications in terms of achieving a more accurate IOL calculation for cataract patients who've had LASIK. We're conducting a study of this application, which is growing more important as LASIK patients age.

Dr. Huang is director of the Laser Vision Center at Doheny Eye Clinic in Los Angeles and associate professor of ophthalmology and biomedical engineering at the University of Southern California.