Beauty may lie in the eye of the beholder but perfection doesn’t — especially in the optical sense. That imperfection makes imaging inside the eye for diagnostic and other purposes a challenge. Now a team from Cardiff University in the UK, the University of Murcia in Spain and the Paris-Meudon Observatory in Meudon, France, has demonstrated a new implementation of adaptive optics that can be applied to imaging of the eye. Team leader Wolfgang Drexler, a Cardiff biomedical imaging professor, noted that key to the approach is the use of a magnetic deformable mirror. The team, with a significant contribution by postdoctoral researcher Enrique J. Fernández, used the mirror to create a correcting device that can cancel out aberrations within a few seconds, potentially enhancing almost any imaging in the eye. “I do see this correcting device applied in any kind of retinal imaging modality — fundus camera, scanning laser ophthalmoscope, optical coherence tomography — to significantly improve image contrast and resolution,” Drexler said. The left column shows a color-coded map of the normal aberrations found in a young eye; red depicts negative values and dark blue, positive ones. The middle left image shows the effect of aberrations on vision, and a retinal image obtained with ultrahigh-resolution optical coherence tomography from the same eye is on the bottom left. In the right column are the effects of correction of monochromatic and chromatic aberration (pancorrection) on the images from the left column. Courtesy of Wolfgang Drexler, University of Cardiff, and Enrique Joshua Fernández, University of Murcia. In the device, the researchers made use of adaptive optics. Originally developed for ground-based astronomy so that stars can be seen clearly through wavering air, adaptive optics employs adjustable optical surfaces to correct for changing optical aberrations. Such deviations from the ideal can be described by Zernike polynomials, and the ability to correct for aberrations can be referred to by the order of the polynomial. By moving parts of a segmented mirror appropriately, astronomers can cancel out the twinkling of stars brought about by the atmosphere. A figure of merit for such efforts is the Strehl ratio, which compares the actual optical performance to that of the ideal. A ratio of 1.00 indicates perfect performance. In ophthalmic applications, the adaptive optics device of choice has been a deformable membrane mirror, but there are problems with the current technology. For one thing, the mirrors have trouble correcting higher order aberrations because of the limited stroke of existing devices. Another problem has been cost. Adaptive optics technology has been too expensive for mass use in commercial ophthalmoscopes. Two years ago Drexler became aware of a magnetic deformable mirror from Imagine Eyes of Orsay, France. The mirror consists of a flexible, silver-coated membrane driven by 52 individually addressable magnetic actuators. When a voltage is applied to the actuators, they attract or repel magnets attached to the back of the membrane, thereby pushing or pulling the membrane over a 17-mm-diameter area. The mirror offers up to sixth-order Zernike correction with no hysteresis and a highly linear response to a driving voltage. The mirror’s features made it attractive for the research in which Drexler and the others were engaged. “We are combining adaptive optics with optical coherence tomography [OCT],” he said. He noted that OCT uses broadband light sources. For that reason, chromatic aberration can present a problem when light of different wavelengths separates in the process of traversing an optical path. In the case of retinal imaging, chromatic aberration of the cornea and lens limits the axial resolution. It also reduces the contrast of the point spread function in the retina, adversely affecting transverse resolution. To minimize chromatic aberration and maximize resolution, the researchers designed an achromatizing lens. They constructed it from two different glasses, with one sandwiching the other in a flint glass-crown glass-flint glass configuration specifically intended for the 700- to 900-nm near-infrared spectral range. Combined with the magnetic deformable mirror, this allowed them to achieve a more complete correction of monochromatic and chromatic aberrations. As detailed in the Oct. 2 issue of Optics Express, they used a helium-neon laser from Melles Griot operating at 632.8 nm for alignment in their demonstration setup. They used a Ti:sapphire laser from Femtolaser Productions GmbH of Vienna, Austria, operating at 800 nm with a 140-nm optical bandwidth to illuminate a subject’s retina after sending the beam through 100 m of fiber for safety reasons. They collected the light exiting the eye from this illumination and sent it into a Hartmann-Shack wavefront sensor from Imagine Eyes. The sensor measured the aberrations of the subject’s eyes and controlled the deformable mirror to correct for those imperfections under the direction of software supplied by the company. With this setup, the researchers attempted to correct the aberrations in the eyes of five subjects. In the case of four, the Strehl ratio was 0.01 to 0.03 at the beginning of the procedure for the 800-nm center of the spectral range of interest. In less than two seconds of device operation, the Strehl ratio was above 0.86 in all cases and above 0.94 in three of them. The fifth subject was quite nearsighted and also had an irregular cornea that introduced extreme high-order aberrations out well past the sixth order. In this case, the system corrected the Strehl ratio from 0.05 to 0.7, a performance that exceeded Drexler’s expectations. “I was extremely surprised that this correcting device can correct extremely high order aberrations in addition to high myopia,” he said. Being able to handle higher-order aberrations could be important in elderly patients because their eyes are more likely to exhibit such characteristics. The capability also means better imaging spot size and contrast, Drexler noted. He added that this capability means that there is no need for the addition of lenses for this task, as is the case with other adaptive optics implementations. That, in turn, could mean a lower-cost solution. It might be possible to use this adaptive optics technology someday for vision correction, but Drexler sees it being used primarily in ophthalmic imaging and related diagnostics applications. A key step will be proving the clinical significance of an adaptive optics-based imaging system, which will require clinical studies and gathering data on a wider sample of eyes. As for turning the setup into a product, Drexler said that it would require some careful consideration of the various options for the first ophthalmic diagnostic device based on this adaptive optics implementation. “It is critical to choose the proper imaging technique to combine with AO in terms of commercialization,” he said.