OCT INSIGHTS

The Future of Commercial OCT?

Clinical applications of advanced imaging technology

Brandon J. Lujan, MD

The future of commercial optical coherence tomography was discussed recently at the retina subspecialty day of the AAO meeting in Chicago. The most easily realized of the improvements offered there fit the general categories of the technological hurdles that were overcome in the evolution from time-domain to spectral-domain OCT: light sources and speed. Research devices have already incorporated many of these advanced capabilities into expensive in-house hardware systems that serve as a test bed for advanced development.

SDOCT systems accomplish their improvement in axial resolution over older OCT systems primarily by the increased bandwidth of their super-luminescent diode (SLD) light sources. This increase in bandwidth has allowed an improved ability to detect the precise depth of the source of reflected light from within the retina. This has translated into an increased axial resolution and the ability to discern individual retinal layers, particularly in the outer retina.

While SLDs previously only existed on expensive custom-built laboratory systems due to their expense, commercialization became possible as costs were driven down by increased use of SLDs in the telecommunications industry. This trend will continue and bandwidths likely will continue to increase.

The use of light sources that are centered at a longer wavelength, near 1 µm, are also an area of possible commercial innovation. Longer wavelengths offer the promise of improved visualization of choroidal structure.

It is intriguing to think that secrets to macular disease, as well as answers to clinical questions that affect management decisions, will be found in the choroid, but this has not (yet) been borne out. While an increased understanding of biology and the descriptions of the choroid in different types of macular pathology are scientifically interesting, the widespread commercial viability of such systems in the absence of a true, driving clinical need would be a fairly large gamble for any company to invest in beyond the value derived from having something new on the market.

In SDOCT, the two-orders-of-magnitude speed increase over time-domain OCT has been obtained by eliminating the rate-limiting step of a moving reference mirror to determine depth information. This has been replaced by the use of a spectrometer and by mathematical transformations that analyze the entire reflected signal from one retinal position simultaneously. This was truly a major breakthrough in OCT technology because the increased speed has allowed for a much denser scan sampling, such that the space between individual crosssectional scans is very unlikely to containclinically relevant macular path ology. The prospect that an even higher scan density would allow the OCT to be even more sensitive to retinal pathology is intriguing, but can only be answered as faster devices are compared with existing SDOCT systems in eyes containing pathology, and this has not been done yet.

Independent of the increased scan density, the elimination of eye motion is indeed an exciting prospect for the next generation of OCT systems. The incorporation of an eye tracker into existing SDOCT systems is helpful to obtain small amounts of data from specific locations, but this is quite impractical to use to obtain a dense scan pattern.

A better solution would clearly be to have a system capable of such speed that the eye cannot possibly have considerable motion during the brief pulse of a scan. This could allow not just a more accurate and reproducible quantitation of topographic disease features, but it could also allow speckle reduction across a dense retinal volume. If this could be accomplished without a drop-off in signal quality and axial resolution and without a substantial increase in expense to manufacture, then this is the most likely direction that future commercial systems will go.

Spectral-domain OCT systems represent a major advance in OCT technology and are by no means close to obsolescence. In fact, the magnitude of the improvements in this technology compared to previous time-domain systems dwarfs any new potential innovations suggested at the AAO and in this article. Indeed, the technology of OCT far precedes the science. RP