Using two deformable mirrors and a customized achromatizing lens, researchers at the University of California, Davis, have greatly improved the resolution of ophthalmic optical coherence tomography (OCT), a mainstay in the clinical treatment of retinal and optic nerve diseases. The instrument, which is only in the research stages at present, could offer clinicians the ability to detect and treat eye disease long before they can with existing technology.Researchers scanned the area of the retina of a 35-year-old volunteer with a more traditional adaptive optics OCT system with 6.5-μm axial resolution (a, left) and with an ultrahigh-resolution OCT system with 3.5-μm axial resolution (a, right). The adaptive optics focus was set on photoreceptor layers. Dashed rectangles mark the photoreceptor layers in both images, which are magnified twofold in b and c, respectively. The abbreviations represent the other retinal layers. RNFL = retinal nerve fiber layer; GCL = ganglion cell layer; IPL = inner plexiform layer; INL = inner nuclear layer; OPL = outer plexiform layer; ONL = fibers of Henle with outer nuclear layer; OLM = outer limiting membrane; I/OS = inner/outer segment junction; VM = Verhoeff’s membrane; RPE = retinal pigment epithelium; Ch = choriocapillaris and choroid. Reprinted with permission of Optics Express.OCT operates like an interferometer, except that it uses the data to create an image of the target area. The technique has proved extremely useful in ophthalmic imaging; however, the eye’s optics have limited its ability to deliver images with high lateral resolution. These problems are almost entirely the result of natural imperfections that occur in the eye’s “refractive interfaces,” explained Robert J. Zawadzki, lead author of a paper on the new technique published in the May 26, 2008 issue of Optics Express. These imperfections create high-order aberrations as well as low-order ones such as defocus and astigmatism. Additionally, the eye’s tissue (mainly water) produces chromatic aberrations. All these reduce the quality of OCT images.Zawadzki explained that most approaches to correcting the aberrations for OCT begin with using lenses to fix the lower-order types. “There is no currently available deformable mirror capable of compensating aberrations fully for the range of aberrations found among the normal population,” he explained. “Some investigators use a single deformable mirror after correcting lower-order aberrations with trial lenses placed in front of the eye or special optical systems with moving elements. This does not work well with OCT because it creates an imbalance in dispersion or path length difference between the reference and sample arms.”Zawadzki and his colleagues opted instead to use a second deformable mirror. “Correction of lower-order aberrations with a deformable mirror having a large stroke but a limited number of actuators is straightforward to implement with OCT,” he said. Then once the lower-order aberrations are compensated, the full dynamic range of a second deformable mirror with lower stroke and more actuators can be used for higher-order aberration correction.With most higher-order aberrations corrected with a second deformable mirror, the group turned its attention to cleaning up the effects of longitudinal chromatic and transverse chromatic aberrations, which are both tied to the effect that chromatic aberrations have on image formation on the retina. When longitudinal chromatic aberration is present, images of different wavelengths focus at different depths, creating a slightly out-of-focus image. When transverse chromatic aberration is present, images of different colors are shifted to the right or left or up or down. “Visually, this can be observed as color fringes at the edge between white and black,” Zawadzki said.Because longitudinal chromatic aberration is wavelength-dependent, the researchers could compensate for it with a specially designed achromatic lens. Transverse chromatic aberration, however, is more complex to correct and would require more advanced lens designs, he said. Thus, they wanted to quantify this type of aberration and describe ways to minimize it with their current system.Alignment is keyThe researchers discovered that transverse chromatic aberration varies greatly, depending upon the camera angle, the angle of the eye and the size of the pupil. Using a reduced eye model, they found that keeping the eye’s nodal point on the axis of the camera eliminated this type of aberration. Although knowing the nodal point in a human eye is impossible, the information they learned allowed them to concentrate on other potential solutions for reducing transverse chromatic aberration without resorting to additional optics. Clinical OCT imaging is typically done with the subject’s head positioned on a chin rest, but the researchers tested their OCT system with a forehead rest and bite bar on a movable stage. This setup typically allows positioning of the pupil to within 0.5 mm. They note that many clinical devices do not require such accurate eye pupil alignment. The researchers tested the setup on a number of people and compared the results with OCT images using only adaptive optics for image correction. The first test compared images from an older system using an SLD 371-HP light source and the second using a Broadlighter SLD T840-HP, both from Superlum of Carrigtwohill, Ireland. The increased spectral bandwidth and, therefore, axial resolution available from the newer light source are clearly visible in the images. Using adaptive optics in optical coherence tomography enables researchers to improve the clarity of images of the layers of the retina. A traditional fundus photo provides very little detailed information about the health of a particular region of the retina. With adaptive optics, the OCT volume acquired over a 0.25 × 0.3-mm retinal region can be reconstructed into the individual retinal layers: (from top to bottom) the nerve fiber layer (NFL), ganglion cell layer (GCL), outer plexiform layer (OPL) and outer segment layer of photoreceptors (OS). Courtesy of Robert J. Zawadzki.Likewise they compared their new system with and without using the achromatizing lens. The results were equally clear. The achromatizing lens added such detail to images of the photoreceptor layers that the researchers say that it may be possible to use the system to measure the length of individual photoreceptor inner and outer segments. Lastly, to increase image contrast, the researchers also employed a speckle reduction system that averaged multiple frames. In the tests, the researchers resolved microscopic detail in clinical diagnostic cases. In one case, they produced an image showing microtraction, in which a portion of the retina has torn. The microtraction was too small to be seen either with a commercial OCT device or with fundus photography. In addition, they produced images that diagnosed a small blind spot near a patient’s fovea that was neither explained with other diagnostic tests nor viewable using other modalities.Zawadzki said that, although in OCT it is the transverse resolution that needs the most improvement, axial resolution improvement is theoretically possible. He added that his group is exploring the system’s possible applications, especially in the area of imaging retinal structures at the cellular level. “To date, we have used this and our previous instruments to image photoreceptor dystrophies, optic neuropathies and cases of unexplained vision loss,” he said. Besides these clinical applications, the researchers are planning to implement this system in experiments that would provide insight into normal development and aging of the visual system.