We studied the long-term optical performance of an adaptive optics scanning laser ophthalmoscope that uses a liquid crystal on silicon spatial light modulator to correct ocular aberrations. The system achieved good compensation of aberrations while acquiring images of fine retinal structures, excepting during sudden eye movements. The residual wavefront aberrations collected over several minutes in several situations were statistically analyzed. The mean values of the root-mean-square residual wavefront errors were 23-30 nm, and for around 91-94% of the effective time the errors were below the Marechal criterion for diffraction limited imaging. The ability to axially shift the imaging plane to different retinal depths was also demonstrated.
© 2011 OSA
Adaptive optics (AO) has been integrated into ophthalmic instruments to obtain high-resolution images of living human retinas. A typical AO system uses a wavefront sensor (WFS) to sense aberrations, which are then corrected using a wavefront corrector (WFC) in a closed-loop manner. AO-equipped instruments, such as AO flood illumination fundus cameras [1–3], AO scanning laser ophthalmoscopes (SLOs) [4–6], and AO optical coherent tomography (OCT) equipment [7–9], have or will become indispensable to studying the structural and functional aspects of vision and retinal pathologies [10–12]. A key component of such systems is the wavefront corrector, which generally can be classified into two types: deformable mirrors (DMs) and liquid crystal spatial light modulators (SLMs). The DMs have physically deformable surfaces, whereas the SLMs have electronically controlled refractive indexes.
The pioneer research on AO retinal imaging conducted by Liang et al. used a bimorph DM with 37 actuators to correct ocular aberrations . Recently micro-electro-mechanical system (MEMS) DMs and magnetic DMs have been widely utilized [13–17]. However, one problem with these DMs is the small number of actuators, which usually ranges from a few tens to a few hundred. Noble et al. studied the relationship between the number of actuators and the achievable Strehl ratio when compensation for the aberrations in human eyes . One drawback of DMs with a small number of actuators is that there is a trade-off between the stroke and the precision of wavefront correction. To increase the dynamic range and improve the precision of wavefront correction, woofer-tweeter dual DM approaches have been proposed and examined [19–24]. This approach, however, complicates the system design and AO control and also increases the system cost.
Liquid crystal spatial light modulators (SLMs) are another type of wavefront corrector [3,9,25–29]. SLMs are attractive because of their low cost and high density of controllable actuators. SLMs can have hundreds of thousands or even millions of pixels. The effect that SLMs have on the incoming wavefront is similar to an array of high-density segmented mirrors, allowing crosstalk between pixels to be ignored. Moreover, they are easy to use because they have no moving parts. This feature makes them suitable for use in harsh environments. However, SLMs also have some shortcomings, such as polarization dependence, limited wavelength bandwidth, and low light-utilization efficiency. The low light utilization efficiency is a fatal drawback for ophthalmic systems because the reflected light signal returning from the retina is very weak. However, the problem can be greatly reduced by adopting up-to-date liquid crystal on silicon (LCOS)-SLM techniques . An LCOS-SLM that has a large fill factor and a multilayer mirror to achieve higher light-utilization efficiency (more than 90% for designed wavelengths) is commercially available .
This LCOS-SLM has been used in a retinal imaging system that utilizes a superluminescent diode (SLD) with a central wavelength of 780 nm as an AO beacon and a 655 nm pulsed laser diode for retinal imaging . The system obtained images of photoreceptors of healthy eyes even with broadband light after AO correction. This LCOS-SLM has also been used in a two-source AO-SLO system to correct the aberrations in living eyes . With this AO-SLO system, we have successfully acquired high-quality images of retinal tissue of hundreds of patients’ eyes, including patients with diabetic retinopathy, macular holes, central serous chorioretinopathy, and abnormalities in microcirculation [12,33]. However, the AO performance of the system has not been fully reported so far.
As a metric to evaluate the performance of AO correction, the root-mean-square (RMS) value of the residual wavefront error is frequently adopted. As far as we know, in most AO retinal imaging systems, including the dual DM AO ones, the RMS values were reported to be around 20 nm to 120 nm, yielding Strehl ratios ranging from 0.45 to 0.98 [16,18–23]. “Good”, that is to say, almost diffraction limited, imaging requires that the RMS wavefront error should be controlled to less than λ/14 , which is equivalent to 60 nm at a wavelength, λ, of 840 nm. It is worth noting that the RMS values mentioned above were evaluated within short periods of time, during which eye movements such as eye blinks and saccades did not occur. In practice, “good” AO correction should be maintained for several minutes or more, during which eye movements are unavoidable.
In this paper, we study the long-term aberration correction performance of an AO-SLO prototype consisting of a single LCOS-SLM. The distributions of RMS values collected during several minutes were statistically analyzed. We also demonstrated an axial focusing capability which allows imaging of tissue at different retinal depths using the LCOS-SLM. Section 2 briefly describes the optical system and imaging method, and Section 3 presents the results achieved in healthy subjects’ eyes.
2. Optical System and Imaging Method
2.1 Optical System
Figure 1 is a block diagram of the AO-SLO system used in this study. The system mainly uses an LCOS-SLM to correct aberrations that arise primarily from the cornea, crystalline lens, and tear film of the subject eye, a Shack–Hartman (SH) WFS to measure the aberrations, raster scanning optics to two-dimensionally scan a flying spot at the retina, and an avalanche photodiode (APD) to detect the light reflected back from this spot. The system uses two different light sources, one as an AO beacon and one for SLO imaging.
The LCOS-SLM (Hamamatsu X10486) is a reflective phase-only modulation device equipped with a monolithic silicon circuit for electrically controlling the orientation of parallel-aligned nematic liquid crystal molecules and a multi-layer dielectric mirror for enhancing the light utilization efficiency. The number of pixels and the pixel size are 792 × 600 and 20 μm × 20 μm, respectively. The gap between pixels is 0.4 μm. The phase modulation range is almost 2π radians. However, an effective stroke of more than 20 wavelengths is achieved by using a phase wrapping technique . The phase is accurately controlled by using an 8-bit digital signal supplied via a digital video interface. Figure 2 shows a photograph of the LCOS-SLM and a typical curve of phase modulation versus input signal level. The good linearity between the input and output indicates that a wavefront can be reproduced with great accuracy.
The WFS consists of a high-speed intelligent vision sensor  and a square lenslet array. The WFS has 25 × 25 sampling spots in a 10.0 mm × 10.0 mm active area. The scanning optics includes a resonant scanner (SC-30, Electro-Optical Products Corporation, NY) for rapid horizontal scanning (HS) and a galvanometric scanner (6210H, Cambridge Technology, MA) for vertical scanning (VS).
The LCOS-SLM, the lenslets of the WFS, the resonant scanner, and the galvanometer are designed and carefully aligned to be conjugate to each other. Each component is placed at the image plane of an afocal telescope consisting of spherical mirrors. The use of spherical mirrors instead of lenses in the afocal telescope design has the benefit of reducing back-reflections and eliminating chromatic aberrations in the system, which is important because the system operates at multiple wavelengths. One of the issues in using the reflective spherical mirrors is that the off-axis configuration introduces astigmatism. This problem was solved by optimizing the system design. The pupil of the subject eye is also aligned at the conjugate plane of the LCOS-SLM and the WFS. The magnification of the pupil at the LCOS-SLM and the WFS is 1.4. According to the active areas of the WFS and LCOS-SLM, a maximum diameter of 7 mm at the pupil of the eye can be measured. However, in this study, we did not use the full areas of the devices. The diameters of the light beams are 5 mm at the eye’s pupil, and 7.2 mm at the planes of the LCOS-SLM and WFS. The expected spot sizes on the retina are approximately 3 μm.
AO beacon light emitted from the 780 nm laser diode (LD) passes through the raster scanning optics and then illuminates the retina of the subject’s eye. The light scattered from the retina passes back through the scanning optics, is reflected from the LCOS-SLM, and finally reaches the WFS. The AO controller, which includes a personal computer and custom software, reads the digitized image from the WFS camera, calculates wavefront aberrations and correction patterns, and then outputs the patterns to the LCOS-SLM. This process is continued in a closed-loop at a rate of 10Hz.
Meanwhile, the retinal imaging light from the superluminescent diode (SLD) with a bandwidth of 50 nm and center wavelength of 840 nm also illuminates a small area of the retina with a flying spot after passing through the scanning optics. The light reflected back from the spot is de-scanned by the scanning optics and then reaches the LCOS-SLM. The light reflected from the LCOS-SLM is corrected and relayed to a confocal detector consisting of a precisely aligned aperture and a fast APD. The output of the APD, when synchronized with the scanners, produces a corrected image that reflects the changes in reflectivity of the scanned area in the retina. The field of view of the instrument is 1.5 deg × 1.5 deg.
To simplify the operation and also to enhance the practical usefulness of the AO-SLO instrument, the system also integrated a pupil camera for aligning the pupil, a Badal stage for compensating for the major refractive error of the eye, and a wide-field-of-view (WFV) retinal imager, which has a field of view of 28 deg × 24 deg, for monitoring and identifying the region to be imaged. The AO system, high-resolution SLO imager, and WFV imager were independently controlled in this study.
2.2 System Implementation
The developed AO-SLO prototype was clinically tested at Kyoto University Hospital. Three healthy subjects participated in this study (they are noted here as HH, IT, and OT). All subjects had myopic refractive errors ranging from 2 D to 6.5 D, and astigmatic refractive errors ranging from 0.25 D to 0.75 D. The subjects’ eyes were tested non-mydriatically, but they had pupil diameters of at least 5 mm. The room was darkened to naturally dilate the pupil, as well as to reduce background illumination. The optical powers for the AO beacon and the high-resolution SLO imaging were 70 μW and 210 μW, respectively, at the cornea. The total power was lower than the ANSI safety limits . Human testing was performed with approved informed consent given by all subjects.
During measurement, the subject put his or her head on a chin-and-forehead rest and fixated at a distant target. The operator adjusted the chin-and-forehead rest and the position of the Badal stage if needed. At the beginning of imaging, the operator aligned the subject’s eye by adjusting the three-dimensional translation stages of the chin-and-forehead rest so that the pupil was at the center of the viewing field of the pupil camera and then activated the WFS in the sensing-only mode and adjusted the Badal stage to the best focusing position. This pre-adjustment could be completed within half a minute. Once the pupil and Badal stage were positioned properly, the operator clicked control buttons to start the AO correction mode and also to record SLO images, which are real-time videos at 10 Hz. The operator could adjust the focus and eccentric position to the depth and the position of interest on the retina. The time for testing one eye, including pre-adjustment, was about 3–6 minutes.
The wavefront aberrations measured by the WFS are represented by the coefficients of Zernike polynomials, which is a series of orthonormal polynomials defined on a unit circle. The AO controller calculated the correction pattern based on the measured aberrations using a closed-loop control algorithm and applied the correction pattern to the LCOS-SLM. We handled the defocus term specially so that it could be included or excluded in the correction pattern. If the correction pattern excluded the defocus, then the defocus was uncorrected by the AO loop, and correspondingly the evaluation of the performance of AO correction did not include the defocus. On the other hard, if the correction pattern included the defocus, the defocus was also corrected by the AO loop, and the evaluation values included the defocus. Additional amounts were added to the defocus term of the Zernike aberrations to axially shift the imaging plane to acquire images at different retinal depths. These options for system implementation increase the flexibility and usefulness in practical applications.
3.1 Imaging with and without Adaptive Optics Correction
Figure 3 shows the retinal images acquired from three healthy subjects with and without aberration correction. The left column (Figs. 3(a), 3(c) and 3(e)) shows the images without AO correction for subjects HH, IT, and OT, respectively, whereas the right column (Figs. 3(b), 3(d) and 3(f)) shows the corresponding results with AO correction. All images are single frames extracted from original recorded videos. The image distortion generated by the sinusoidal scan pattern of the resonant scanner was compensated. The real field of view was approximately 1.3 deg × 1.5 deg, corresponding to an area of approximately 400 μm × 450 μm on the retina. Photoreceptors (bright spots) can be resolved in the corrected images (Figs. 3(b), 3(d) and 3(f)), whereas the uncorrected images are just blurred (Figs. 3(a), 3(c) and 3(e)). The spatial resolution and contrast were significantly improved after AO correction.
Figure 4 shows the wavefront maps with and without AO correction for the above three subjects. The left column (Figs. 4(a), 4(c) and 4(e)) shows the uncorrected wavefront maps, whereas the right column (Figs. 4(b), 4(d) and 4(f)) shows residual maps after AO correction. The fringes in the maps are due to 2π phase wrapping, and one fringe represents a 2π phase retardation, equivalent to a wavefront error of one wavelength. Therefore, the more fringes, the larger the wavefront error. We can see several fringes in the uncorrected wavefront maps (Figs. 4(a), 4(c) and 4(f)), whereas the maps with AO correction are almost flat. The RMS wavefront errors of the uncorrected wavefronts were 0.35 μm, 0.94 μm, and 1.17 μm, which were reduced to 0.020 μm, 0.024 μm, and 0.026 μm, respectively, after AO correction. The peak-to-valley (PV) values were also significantly reduced after AO correction.
Figure 5 shows the aberrations versus Zernike mode for subject IT with and without AO correction. The horizontal axis is the index of the Zernike mode. Index 3 is defocus, 4 and 5 are astigmatism, 6 and 7 are coma, 8 and 9 are trefoil, 10 is spherical aberration, 11 and 12 are secondary astigmatism, 13 and 14 are quadrafoil, and 15-20 are the 5-th order aberrations. Note that the magnitudes with AO correction are enlarged 100 times in order to show them at the same scale. The subject’s eye had large astigmatism of about 0.9 μm, as indicated in the graph. Both lower- and higher-order aberrations were dramatically reduced with AO correction.
Figure 6 shows the time course of the RMS wavefront error measured while imaging the subject’s retina. At approximately 5 seconds, the operator toggled a switch for changing from the WFS sensing-only mode to the AO correction mode. Once the AO loop was turned on, we observed a sudden drop in RMS value. The RMS value approached the average value of 24 nm within 1 second after turning on the AO loop. For a one-second period during which aberrations were well-corrected, the mean RMS wavefront error was only 12 nm, and the corresponding Strehl ratio was 0.99 according to the Marechal approximation . The control software was optimized by consideration of the temporal characteristics of the devices used. The slowest device was the LCOS-SLM and its response time was approximately 50 ms. As a result of the optimization, the system can respond quickly to sudden and large changes in aberration as shown in Fig. 6.
We observed “V”-shaped fluctuations in the RMS value, as indicated in the inset of Fig. 6. The sharp fluctuations were caused by blinking of the eye and associated saccadic movements. Because the eye’s pupil aperture was closed during the blinks, aberration measurement and also retinal imaging were not possible at these times. This was confirmed by retinal images. Figure 7 shows a series of images captured during the course of the movements. The time stamps are the same as the x-axis in Fig. 6. Figure 7(a) is an image one frame ahead of the image acquired during blinking, shown in Fig. 7(b), which shows nothing because the eye was closed. Figure 7(c) shows the image distorted by a saccade when recovering from the blink. Figure 7(d) is the image acquired after the eye movements. Comparing those images, we found that the images showed some transverse shift before and after the movement; however, the image quality was almost the same, indicating that the AO loop operated properly before and after the movements. We used the blink frames for timing registration between the retinal images and the AO data.
3.2 Long-Term Performance of Adaptive Optics Correction
The long-term stability of AO correction was demonstrated. Retinal image acquisition and AO data collection were performed simultaneously for several minutes. Figure 8 shows an example of the time course of the RMS wavefront error achieved when imaging the retinal blood vessels of subject IT. The level of RMS wavefront error without AO correction is also shown by the red dashed line. We found that for most of the time the RMS wavefront error was significantly reduced by AO correction. Of the entire 180 s observation time, correction was inadequate for a total period of about 4 s because of the eye blinks, as indicated by the spikes in the graph. The average RMS value over the effective time, which was about 176 s, was 30 nm, compared with about 900 nm without AO correction. In Fig. 8, the RMS values during the blinks were non-zero (about 450 nm in this graph), in contrast to that shown in Fig. 6 where the zeroed RMS values were observed during the blinks. This difference comes from the fact that the position of the Badal stage was set at the best focus position on the photoreceptor layer, and we added a constant bias to the defocus term of AO correction aberrations in order to move the imaging plane from the photoreceptor layer to the blood vessel layer. The non-zero values during blinks were exactly equal to the additional bias. In Fig. 8, besides the large spike-like fluctuations, we can also see many small fluctuations. These small fluctuations were caused by eye movements such as saccades and drifting. Note that the average over the effective time includes the effects of these movements. The residual wavefront error was even smaller when the effect of these movements was excluded. For example, the average RMS value estimated over a 50-second blink-free period (i.e., the period from 45 s to 95 s in Fig. 8) was 25 nm, and the standard deviation was 16 nm. Furthermore, within a selected 2 s period of “good” fixation, during which neither blinking nor saccade movements occurred, the RMS value was estimated to be only 11.5 nm with a standard deviation of 4.2 nm.
Figure 9 shows the distribution histogram and cumulative probability curve of the RMS values. The histogram (Fig. 9(a)) shows a non-Gaussian shape. The long tail of the distribution represents the RMS values estimated during eye movements. The modal value was 18 nm, and the median value was 23 nm. In Fig. 9(b), the vertical broken line indicates an RMS value of 60 nm, which is the wavefront aberration corresponding to the Marechal criterion for the Rayleigh diffraction limit at a wavelength of 840 nm , and the horizontal broken line shows the corresponding cumulative probability. For 91.4% of the effective time, the RMS value was within the Marechal criterion, showing that almost diffraction-limited imaging was possible for more than 90% of the effective time.
The AO data for subjects HH and OT showed similar statistical characteristics, as shown in Fig. 10 , even though they were acquired under different experimental conditions. These conditions were the focusing depth and the eccentric position at the retina, the observation time, and the method of correcting defocus. Plot IT was obtained with subject IT while focusing at the retinal blood vessel layer. The effective observation time was about 176 s. The AO correction included the defocus term, and thus the RMS wavefront error included second- and higher-order aberrations. Plot OT was obtained with subject OT while focusing at the retinal nerve fiber layer, but the scanning area was shifted continuously across the retina. The effective observation time was about 233 s. The AO did not correct the defocus, which was corrected by adjusting the Badal stage. The RMS value did not include the defocus. Plot HH was obtained with subject HH while focusing at the photoreceptor layer near the fovea. The effective observation time was about 80 s. In the first half time (0–40 s), the defocus was corrected by adjusting the Badal stage, and in the latter half (40–80 s), the defocus was corrected by AO. Therefore, half of the measured values included the defocus, and remaining half did not include the defocus.
These simulated conditions may cover the most situations in practical clinical applications. From the experimental data, we found that the mean RMS wavefront errors over the effective observation time were 30 nm, 28 nm, and 23 nm, and the proportions of the times where diffraction-limited imaging was possible were approximately 91.4%, 93.8%, and 92.8% for the three subjects IT, OT, and HH, respectively.
Figure 11 shows an AO-SLO image (and video) of the retinal blood vessel layer, two retinal capillaries were observed. The 10-second long video is a part of the whole 3-minute real-time video acquired in the experiments, and Fig. 8 described above shows the RMS wavefront errors recorded synchronously. We can see the blood flows in the vessels. The frame-to-frame shifts and intra-frame distortions were caused by drifting and saccadic movements of the eye. It is worth mentioning that most frames, excepting those during blinking, were at nearly the best focused position on the blood vessel layer, despite the eye accommodation. The images were stabilized by using a post-processing algorithm, but no frame average was applied. The scale bar represents 100 μm. The frame rate was 10 Hz.
3.3 Axial Scanning in Retinal Depth Direction with LCOS-SLM
We demonstrated the ability to focus at different retinal depths with the LCOS-SLM. This focusing was accomplished by adding a bias to the defocus aberration in the AO correction loop. A demonstration of continuously shifting the imaging plane is shown in Fig. 12 . Figure 12(a) shows the time courses of the added refractive power (A-RP; red dash-dotted line) and the residual refractive power (R-RP; blue solid line). The A-RP and R-RP were calculated from the added bias and the Zernike defocus coefficient measured by the WFS. We found that the detected refractive power of R-RP was consistent with the applied bias of A-RP. Some differences between them, as shown in Fig. 12(a), were due to eye accommodation. The system was initially focused on the photoreceptor layer with zero bias. The applied refractive power (the value of the added bias) was varied from 0 D (0 μm) to 1.73 D (−1.6 μm), and then from 1.73 D (−1.6 μm) to −1 D (0.9 μm) in steps of 0.0086 D (bias value, 0.0078 μm). The above procedure corresponded to axially shifting the imaging plane from the posterior retina up to the anterior retinal surface, and then returning back to posterior retina.
Figure 12(b) shows a series of instantaneous images acquired while varying the bias. The images were captured while focusing at the photoreceptor layer (PRL; panel 1 of Fig. 12(b)), and then moving to the blood vessel layer (BVL; panel 3), the nerve fiber layer (NFL; panel 4), beyond the NFL (panel 5), and then returning back to the NFL (panel 6), the BVL (panel 7), and the PRL (panel 9). Figure 12(b) also shows images acquired between the PRL and the BVL, where both photoreceptors and blood vessels are recognizable (panels 2 and 8). Each image was a single scan of a 1.3 deg × 1.5 deg area of the retina. The time stamps in the panels of Fig. 12(b) correspond to the x-axis of Fig. 12(a). Comparing the retinal images and the corresponding AO data, we find that “good” focused images of the PRL could be acquired when the bias was −0.05 D to 0.15 D (see the light red bar in Fig. 12(a)), and the ranges for acquiring “good” images of the BVL and NFL were 0.35 D to 0.45 D and 0.52 D to 0.61 D, respectively, as indicated by the light green and blue bars in Fig. 12(a). The results indicate that the focusing control was performed very well.
The range of the above focusing adjustment was 2.73 D, equivalent to axially moving the imaging plane by approximately 1 mm, which is larger than the average thickness of a human retina.
The amount of shifting of the focusing plane can be estimated according to the detected Zernike defocus coefficients. We assume the Emsley’s version of the reduced eye for which in the unaccommodated eye the power is + 60 D and the index of refraction is 1.333 , we estimated the retina thickness from the well-focused PRL to NFL to be approximately 200 μm, which is in agreement with measurements in normal eyes .
We reported the aberration correction performance of an AO-SLO system that uses a high-quality LCOS-SLM as a wavefront corrector and an intelligent vision sensor as the sensing device. The system uses two light sources, one as an AO beacon and one for high-resolution retinal imaging. The AO-SLO system achieved good performance in terms of aberration correction and retinal image acquisition. The long-term stability of AO correction was studied. Good performance was maintained, except during sudden eye movements, such as blinks and saccades. In testing of healthy human subjects, the RMS wavefront errors averaged over several minutes could be reduced to 23-30 nm. The RMS value was even smaller when estimated over good-fixation periods of several seconds. Almost diffraction-limited imaging was possible for more than 90% of the effective time. We believe that the stable long-term performance was mainly due to the ability to precisely reproduce wavefront aberrations at the LCOS-SLM device. The optimized control software also contributed to the stabilized high performance of the AO-SLO system. Good long-term performance is necessary in order to transfer AO instruments from research applications to routine clinical applications. The ability to axially scan the imaging plane in the retinal depth direction with the LCOS-SLM was also demonstrated. Focusing controlling was performed very well. Images of minute retinal structures, such as photoreceptors, blood vessels, and nerve fiber layers, were easily acquired, and the displacements between different imaging planes were estimated. The results indicate that the liquid crystal AO system is capable of high-accuracy and high-stable wavefront manipulation and shows great promise for practical high-resolution retinal imaging instruments.
This research was supported in part by the New Energy and Industrial Technology Development Organization of Japan (FY2005–FY2009). All testing was performed at Kyoto University Hospital in collaboration with Yoshihiko Yamada and Susumu Oshima of Nidek Co., Ltd., and Kouhei Takayama and Nagahisa Yoshimura of Kyoto University.
References and links
1. J. Liang, D. R. Williams, and D. T. Miller, “Supernormal vision and high-resolution retinal imaging through adaptive optics,” J. Opt. Soc. Am. A 14(11), 2884–2892 (1997). [CrossRef]
2. J. Rha, R. S. Jonnal, K. E. Thorn, J. Qu, Y. Zhang, and D. T. Miller, “Adaptive optics flood-illumination camera for high speed retinal imaging,” Opt. Express 14(10), 4552–4569 (2006), http://www.opticsinfobase.org/abstract.cfm?URI=oe-14-10-4552. [CrossRef] [PubMed]
3. P. M. Prieto, E. J. Fernández, S. Manzanera, and P. Artal, “Adaptive optics with a programmable phase modulator: applications in the human eye,” Opt. Express 12(17), 4059–4071 (2004), http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-12-17-4059. [CrossRef] [PubMed]
4. A. Roorda, F. Romero-Borja, W. Donnelly Iii, H. Queener, T. J. Hebert, and M. C. W. Campbell, “Adaptive optics scanning laser ophthalmoscopy,” Opt. Express 10(9), 405–412 (2002), http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-10-9-405. [PubMed]
5. S. A. Burns, R. Tumbar, A. E. Elsner, D. Ferguson, and D. X. Hammer, “Large-field-of-view, modular, stabilized, adaptive-optics-based scanning laser ophthalmoscope,” J. Opt. Soc. Am. A 24(5), 1313–1326 (2007). [CrossRef] [PubMed]
6. D. X. Hammer, R. D. Ferguson, C. E. Bigelow, N. V. Iftimia, T. E. Ustun, and S. A. Burns, “Adaptive optics scanning laser ophthalmoscope for stabilized retinal imaging,” Opt. Express 14(8), 3354–3367 (2006), http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-14-8-3354. [CrossRef] [PubMed]
7. B. Hermann, E. J. Fernández, A. Unterhuber, H. Sattmann, A. F. Fercher, W. Drexler, P. M. Prieto, and P. Artal, “Adaptive-optics ultrahigh-resolution optical coherence tomography,” Opt. Lett. 29(18), 2142–2144 (2004). [CrossRef] [PubMed]
8. Y. Zhang, J. Rha, R. Jonnal, and D. Miller, “Adaptive optics parallel spectral domain optical coherence tomography for imaging the living retina,” Opt. Express 13(12), 4792–4811 (2005), http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-13-12-4792. [CrossRef] [PubMed]
9. E. J. Fernández, B. Povazay, B. Hermann, A. Unterhuber, H. Sattmann, P. M. Prieto, R. Leitgeb, P. Ahnelt, P. Artal, and W. Drexler, “Three-dimensional adaptive optics ultrahigh-resolution optical coherence tomography using a liquid crystal spatial light modulator,” Vision Res. 45(28), 3432–3444 (2005). [CrossRef] [PubMed]
11. D. X. Hammer, N. V. Iftimia, R. D. Ferguson, C. E. Bigelow, T. E. Ustun, A. M. Barnaby, and A. B. Fulton, “Foveal fine structure in retinopathy of prematurity: an adaptive optics Fourier domain optical coherence tomography study,” Invest. Ophthalmol. Vis. Sci. 49(5), 2061–2070 (2008). [CrossRef] [PubMed]
12. S. Ooto, M. Hangai, A. Sakamoto, A. Tsujikawa, K. Yamashiro, Y. Ojima, Y. Yamada, H. Mukai, S. Oshima, T. Inoue, and N. Yoshimura, “High-resolution imaging of resolved central serous chorioretinopathy using adaptive optics scanning laser ophthalmoscopy,” Ophthalmology 117(9), 1800–1809 (2010). [CrossRef] [PubMed]
13. N. Doble, G. Yoon, L. Chen, P. Bierden, B. Singer, S. Olivier, and D. R. Williams, “Use of a microelectromechanical mirror for adaptive optics in the human eye,” Opt. Lett. 27(17), 1537–1539 (2002). [CrossRef] [PubMed]
15. C. E. Bigelow, N. V. Iftimia, R. D. Ferguson, T. E. Ustun, B. Bloom, and D. X. Hammer, “Compact multimodal adaptive-optics spectral-domain optical coherence tomography instrument for retinal image,” J. Opt. Soc. Am. A 24(5), 1327–1336 (2007). [CrossRef]
16. E. J. Fernandez, L. Vabre, B. Hermann, A. Unterhuber, B. Povazay, and W. Drexler, “Adaptive optics with a magnetic deformable mirror: applications in the human eye,” Opt. Express 14(20), 8900–8917 (2006), http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-14-20-8900. [CrossRef] [PubMed]
17. M. Mujat, R. D. Ferguson, N. Iftimia, and D. X. Hammer, “Compact adaptive optics line scanning ophthalmoscope,” Opt. Express 17(12), 10242–10258 (2009), http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-17-12-10242. [CrossRef] [PubMed]
18. N. Doble, D. T. Miller, G. Yoon, and D. R. Williams, “Requirements for discrete actuator and segmented wavefront correctors for aberration compensation in two large populations of human eyes,” Appl. Opt. 46(20), 4501–4514 (2007). [CrossRef] [PubMed]
19. D. C. Chen, S. M. Jones, D. A. Silva, and S. S. Olivier, “High-resolution adaptive optics scanning laser ophthalmoscope with dual deformable mirrors,” J. Opt. Soc. Am. A 24(5), 1305–1312 (2007). [CrossRef] [PubMed]
20. M. Mujat, R. D. Ferguson, A. H. Patel, N. Iftimia, N. Lue, and D. X. Hammer, “High resolution multimodal clinical ophthalmic imaging system,” Opt. Express 18(11), 11607–11621 (2010), http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-18-11-11607. [CrossRef] [PubMed]
21. W. Zou and S. A. Burns, “High-accuracy wavefront control for retinal imaging with Adaptive-Influence-Matrix Adaptive Optics,” Opt. Express 17(22), 20167–20177 (2009), http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-17-22-20167. [CrossRef] [PubMed]
23. J. W. Evans, R. J. Zawadzki, S. M. Jones, S. S. Olivier, and J. S. Werner, “Error budget analysis for an adaptive optics optical coherence tomography system,” Opt. Express 17(16), 13768–13784 (2009), http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-17-16-13768. [CrossRef] [PubMed]
24. C. Li, N. Sredar, K. M. Ivers, H. Queener, and J. Porter, “A correction algorithm to simultaneously control dual deformable mirrors in a woofer-tweeter adaptive optics system,” Opt. Express 18(16), 16671–16684 (2010), http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-18-16-16671. [CrossRef] [PubMed]
26. F. Vargas-Martín, P. M. Prieto, and P. Artal, “Correction of the aberrations in the human eye with a liquid-crystal spatial light modulator: limits to performance,” J. Opt. Soc. Am. A 15(9), 2552–2562 (1998). [CrossRef] [PubMed]
27. H. Huang, T. Inoue, and T. Hara, “An adaptive wavefront control system using a high-resolution liquid crystal spatial light modulator,” Proc. SPIE 5639, 129–137 (2004). [CrossRef]
28. T. Yamaguchi, N. Nakazawa, K. Bessho, Y. Kitaguchi, N. Maeda, T. Fujikado, and T. Mihashi, “Adaptive optics fundus camera using a liquid crystal phase modulator,” Opt. Rev. 15(3), 173–180 (2008). [CrossRef]
29. T. Inoue, H. Tanaka, N. Fukuchi, M. Takumi, N. Matsumoto, T. Hara, N. Yoshida, Y. Igasaki, and Y. Kobayashi, “LCOS spatial light modulator controlled by 12-bit signals for optical phase-only modulator,” Proc. SPIE 6487, 64870Y, 64870Y-11 (2007). [CrossRef]
31. T. Shirai, K. Takeno, H. Arimoto, and H. Furukawa, “Adaptive optics with a liquid crystal on silicon spatial light modulator and its behavior in retinal imaging,” Jpn. J. Appl. Phys. 48(7), 070213 (2009). [CrossRef]
32. M. Hanebuchi, J. Sasaki, N. Shibata, Y. Fujisaka, Y. Yamada, H. Mukai, and S. Oshima, “Adaptive optics scanning laser ophthalmoscopy for imaging retina using a single liquid crystal spatial light modulator,” In ARVO 2010 Annual meeting, Poster A557, Program NO 2307, Fort Lauderdale, Florida, May 2–6 (2010).
33. S. Ooto, M. Hangai, K. Takayama, A. Sakamoto, A. Tsujikawa, S. Oshima, T. Inoue, and N. Yoshimura, “High-resolution retinal imaging of the photoreceptor layer in epiretinal membrane visualized with adaptive optics scanning laser ophthalmoscopy,” Ophthalmology 118(5), 873–881 (2011). [CrossRef] [PubMed]
34. M. Born and E. Wolf, Principles of Optics, 7th Ed. (Cambridge University Press, 2001), p. 528.
35. H. Huang, T. Inoue, and T. Hara, “Adaptive aberration compensation system using a high-resolution liquid crystal on silicon spatial light modulator,” Proc. SPIE 7156, 71560F (2009).
36. T. Toyoda, N. Mukohzaka, S. Mizuno, Y. Nakabo, and M. Ishikawa, “Column parallel vision system (CPV) for high-speed 2D-image analysis,” Proc. SPIE 4416, 256–259 (2001). [CrossRef]
37. American National Standards Institute, American National Standard for the Safe Use of Lasers, ANSI Z136.1–2000, (American National Standards Institute, 2000).
38. G. Dai, Wavefront Optics for Vision Correction (SPIE Press, 2008), p. 221.
39. D. Landau, E. M. Schneidman, T. Jacobovitz, and Y. Rozenman, “Quantitative in vivo retinal thickness measurements in healthy subjects,” Ophthalmology 104(4), 639–642 (1997). [PubMed]