With the advent of wavefront technology, it has been possible to quantify lower and higher order ocular aberrations and better understand visual performance. Visual function is degraded by several optical phenomena, including wavefront aberrations, light scattering, and diffraction. Wavefront analysis is one method to indirectly estimate vision impairment caused by optical aberrations by quantifying the total and higher order ocular aberrations of the optical system. The real-time measurement of the eye's wavefront aberrations and the dynamic correction of these aberrations have been described using an adaptive optics system.12
Adaptive optics technology contains two basic components: a repetitively capturing wavefront sensor and a deformable mirror, which compensates for the wavefront error in real time. Some of the limitations encountered with previous adaptive optics systems using electrostatic deformable mirrors and liquid crystal spatial light modulators were recently overcome by the electromagnetic technology. The shape of a deformable mirror is generally controlled by an array of actuators. In an electrostatic mirror, the driving force in each actuator is due to the interaction between two elements loaded with adjustable static electrical charges. In an electromagnetic mirror, each actuator uses a miniature metallic coil that, when supplied with an electrical current, exerts a force on a miniature magnet, following the same principle as a conventional loudspeaker. Fern?ndez3 demonstrated that the deformation range of more than 50 µm of an electromagnetic deformable mirror was suitable for the correction of wavefront aberrations in a wide variety of human eyes.
Adaptive optics systems have been used to measure and correct aberrations of the eye in a "closed-loop" for the assessment of visual quality and also to improve the quality of retinal images for testing of cell anatomy and function.2,4,5 Because the eye's aberrations also limit the resolution by which the retina can be imaged and photographed, adaptive optics systems allow for clinical observation and investigation beyond the limits of other ophthalmoscopic instruments.
The impact of higher order aberrations on visual performance and the benefit expected after higher order ocular aberration correction have already been studied.67 Liang et al1 described the visual benefits of correcting the eye's higher order ocular aberrations using the adaptive optics system and successfully improved optical quality and contrast sensitivity. Applegate et al8 measured low and high contrast acuity after generating aberrated chart letters using the CTView system (Sarver & Associates, Carbondale, 111). The authors demonstrated that visual acuity decreased in a linear fashion when root-mean-square (RMS) wavefront error increased above 0.05 µm, with the slopes of the linear fits varying according to the mode. Earlier studies using computer-generated aberrated visual acuity charts and schematic eye models have shown that higher order ocular aberrations degrade the optical quality of the eye.912 However, the current study used a different approach from earlier studies because it actually modified the optics and then assessed the effect of aberrations on visual acuity.
The aim of this study was to measure and quantify the changes in visual acuity induced by individual Zernike modes of various RMS magnitude in wavefront corrected, diffraction limited eyes using a new electromagnetic adaptive optics system.
PATIENTS AND METHODS
The experiments were performed in the right eyes of nine patients who complied with the following inclusion criteria: clear intraocular media, age between 18 and 45 years, no retinal pathology, and measured pupil diameter larger than 5 mm under our testing conditions. The patient group included two women and seven men. Spherical refractive error in the patients' right eyes ranged between -6.50 and +0.50 diopters (D) and cylinder ranged between -0.25 and - 1.00 D. Higher order ocular aberrations measured in natural pupil conditions and numerically analyzed within a 5-mm natural pupil diameter ranged between 0.10 and 0.32 µm, except for two patients previously diagnosed as having keratoconus whose respective right eye ocular higher order aberrations were 0.89 and 1.04 µm.
ADAPTIVE OPTICS SYSTEM
We used a crxl Adaptive Optics Visual Simulator (Imagine Eyes, Orsay, France) to measure, correct, and modify the patient's ocular aberrations. The main components in this instrument are a 1024 lenslets Shack-Hartmann wavefront sensor, an electromagnetic deformable mirror with 52 actuators, and an internal 800X600 pixels microdisplay monitor. Each sensor microlens samples the wavefront error in a square area of 210X210 pm at the eye pupil plane. The spacing between actuators, calculated at the same plane, is 820 µm.
Three different operating procedures are available: (1) wavefront measurement: ocular wavefront aberrations are recorded while the deformable mirror is set to an aberration-free shape; (2) static wavefront correction/generation: user-defined aberrations are applied and maintained constant, independent of wavefront fluctuations that may occur in the eye under testing; and (3) dynamic wavefront correction/generation: user-defined aberrations are applied using a closedloop system that comprises a double-pass of light through the eye, so that the total (eye + device) aberration encountered along the line of sight remains constant. In the current study, we generated static aberrations by combining wavefront measurement and static wavefront correction/generation. We did not use the dynamic procedure.
In the static wavefront correction/generation procedure, the crxl software allows the operator to define a simulated wavefront that includes a compensation of the aberrations previously measured in the patient's eye and additional adjustable Zernike aberrations up to the 4th order. The device then simulates the user-defined wavefront using an internal closed-loop system that sets the mirror surface to the desired shape and maintains it constant. The system thus generates a static wavefront. The pupil size is adjusted by selecting the diameter of an internal aperture. In the current study, this aperture was set to a diameter of 5 mm. Throughout the visual simulation, an eye-tracking system monitors the relative positions of the instrument's optical axis and the patient's pupil. The eye-tracker data are continuously displayed in a software window that helps to maintain the best possible alignment during the test. The patient's head position is maintained using a standard chinrest.
In the same procedure, the software also continuously displays a residual wavefront that monitors the fidelity of the wavefront achieved by the deformable mirror compared to the wavefront predefined by the user. Software algorithms compute the residual wavefront as the RMS difference between the wavefront measured by the Shack-Hartmann sensor and the expected wavefront.
VISUAL ACUITY TEST
We used the Freiburg Acuity Test (FrACT)1314 software to measure the patients' monocular visual acuity through the adaptive optics system. We ran the FrACT software on a separate laptop PC computer connected to the crxl internal microdisplay monitor. The patient viewed the miniature monitor though the deformable mirror and pupil conjugation lenses of the simulator. These optical elements and the monitor were all coaxially aligned in the system. The monitor subtended a field of view of 124 arcmin horizontal by 93 arcmin vertical. Its white background luminance was adjusted to 50 cd/m2. The patient was stimulated with black Landolt C optotypes and responded by indicating the optotype orientation on a numeric keypad. After each response, the FrACT software automatically modified the optotype size according to a "parameter estimation by sequential testing" (PEST) method and randomly selected the new optotype orientation among eight possible directions.
Each visual acuity test ended after 18 optotype presentations. This staircase method provided a rapid psychophysical procedure to estimate visual acuity thresholds. Because the presentations depended only on the patient's response, the examiner's bias was minimized. Each patient submitted to two training visual acuity tests before beginning the actual experiment.
We applied static wavefront modifications to the patients' eyes using the crxl simulator procedures 1 (wavefront measurement) and 2 (static wavefront correction/generation). We performed all experiments in natural eye conditions (eg, without having recourse to mydriatic or cycloplegic drugs). We first assessed the ocular wavefront of each patient's right eye while the device was presenting a fogged distant target that stimulated the relaxation of the eye's accommodation. The measured aberrations were recorded as a set of Zernike coefficients up to the 10th order. We then defined 18 different simulated wavefronts as follows:
* Simulated wavefront #1: no correction, computed as a flat wavefront;
* Simulated wavefront #2 : sphero-cylinder correction, computed as the opposite of the best least-square sphero-cylindrical fit to the measured wavefront;
* Simulated wavefront #3: sphero-cylinder and higher order ocular aberration correction, computed as the opposite of the measured wavefront;
* Simulated wavefronts #4 to #18: computed by adding the opposite of the measured wavefront and single Zernike modes.
Means and Inter-individual Standard Deviations of the Measured Visual Acuity Data (logMAR)
In defining the latter 15 simulated wavefronts, the single Zernike modes successively included defocus Z(2,0), oblique astigmatism Z(2,? 2), vertical coma Z(3,? 1), oblique trefoil Z(3,? 3), and spherical aberration Z(4,0). Each of these five aberration modes was generated in three different positive amounts specified by Zernike coefficients of 0.1, 0.3, and 0.9 pm. The definition and indexing of Zernike polynomials used in the current study comply with the Optical Society of America standards for reporting ocular wavefront aberrations. The 18 testing conditions are listed in the first column of the Table.
We set the artificial pupil in the adaptive optics device to a diameter of 5 mm and configured the software to apply the 18 predefined wavefronts to each eye in a randomized order. The randomization included 4 repetitions of each of the simulated wavefronts #1, #2, and #3. We measured the patient's monocular visual acuity with each simulated wavefront. Throughout the visual acuity tests, we adjusted the alignment of the instrument with respect to the eye pupil using the eyetracking display. We also continuously monitored the variations of the adaptive optics residual wavefront and recorded any RMS value above 100 nm.
Figure 1. Statistics of logarithm of the minimum angle of resolution (logMAR) visual acuity (VA) measured with best sphero-cylinder (SC) and with full adaptive optics (sphero-cylinder and higher order ocular aberration [SC + HOA]) corrections.
Figure 2. Case profiles of logarithm of the minimum angle of resolution (logMAR) visual acuity (VA) measured with best sphero-cylinder (SC) and with full adaptive optics (sphero-cylinder and higher order ocular aberration [SC + HOA]) corrections. The negative slopes indicate that higher order wavefront correction resulted in improved visual acuity in all eyes.
We measured the visual acuity of each patient's right eye in the 18 wavefront configurations of our experiment. The means and inter-individual deviations computed from the resulting visual acuity data are presented in the Table. The visual acuity test precision, evaluated as the intra-patient standard deviation computed from repeated visual acuity measurements, was a mean of 0.05 logarithm of the minimum angle of resolution (logMAR). Throughout the experiments, the eye pupil size remained larger than the artificial pupil diameter (5 mm) in every case. The alignment of the instrument with respect to the eye pupil was maintained with accuracy better than 0.3 mm, except for occasional short-time deviations due to blinks and saccades. The RMS residual error of the internal closed-loop system remained smaller than 80 nm, except with the two eyes that had large higher order ocular aberrations of 0.89 and 1.04 pm, respectively. In these two eyes, the adaptive optics RMS residual error reached up to 150 nm.
The patients' monocular visual acuity was improved from a mean of +0.41 logMAR in uncorrected eyes to a mean of ?0.01 logMAR with sphero-cylinder correction. Correcting higher order ocular aberrations up to the 4th order, in addition to sphere and cylinder, further enhanced visual acuity up to a mean of ?0.09 logMAR. Figure 1 shows the average results obtained with spherocylinder and sphero-cylinder and higher order ocular aberration corrections. The observed difference in visual acuity was found to be statistically significant (Wilcoxon matched-paired test, P<.05). The inter-individual variability of visual acuity was reduced by both corrections, from a standard deviation of 0.56 logMAR in uncorrected eyes down to 0.13 and 0.11 logMAR with sphero-cylinder and sphero-cylinder and higher order ocular aberration corrections, respectively.
The case profiles for the same data, presented in Figure 2, show that sphero-cylinder and higher order ocular aberration correction resulted in visual acuity equal to or better than that of sphero-cylinder correction in every tested eye. Some subjects experienced a visual acuity improvement of more than 0.15 logMAR, whereas the measured change in other subjects was less than 0.02 logMAR. In the two eyes with the highest aberrations (0.89 and 1.04 pm RMS, respectively), the improvement in visual acuity was 0.09 and 0.08 logMAR, respectively. The data from these two eyes correspond to the two upper lines in the graph of Figure 2.
The modifications in the patients' visual acuity measured when simulating individual Zernike aberrations are presented in Figure 3. Using the visual acuity measured with sphero-cylinder and higher order ocular aberration correction as a baseline, we computed the changes in visual acuity due to the introduction of individual Zernike modes. The statistics of Figure 3 differ slightly from those of the Table; this difference is due to the baseline subtraction. In general, the added Zernike aberrations induced losses in visual acuity that became worse as the aberration coefficients increased.
Figure 3. Statistics of the changes in visual acuity induced by the application of individual Zernike aberrations 0.1 (left graph), 0.3 (center graph), and 0.9 (right graph) ??t?. For each patient, the changes in visual acuity were computed by subtracting a baseline visual acuity value, measured with the best possible wavefront correction, from the visual acuity findings obtained while adding individual Zernike aberrations.
Single Zernike aberrations applied with a coefficient of 0.1 pm resulted in small changes in visual acuity ranging between mean values of 0.03 and 0.05 logMAR, except for the spherical aberration that did not result in any noticeable change (Fig 3, left). The same aberration modes generated in amounts of 0.3 pm induced significant losses in visual acuity (Fig 3, center). The mean visual acuity changes from the full-corrected baseline ranged between 0.13 and 0.16 logMAR and appeared to be practically independent of the Zernike mode number. Increasing the aberration coefficient to 0.9 pm led to poorer visual acuity findings, with changes ranging between mean values of 0.23 and 0.65 logMAR (Fig 3, right).
These latter results, generated with high aberration magnitude, showed a strong dependency on the mode number. Defocus and spherical aberration induced the greatest changes in visual acuity with a mean of 0.6 logMAR. Astigmatism and coma induced visual acuity losses ranged between mean values of 0.3 and 0.4 logMAR, whereas the decrease due to trefoil was a mean of 0.23 logMAR. The statistical analysis confirmed that the mode number, the aberration coefficient value, and the interaction between these two factors had significant effects on the change in visual acuity (analysis of variance, P<.05).
In designing the current study, we chose to experiment with static wavefronts rather than with wavefronts dynamically adjusted to real-time changes in ocular aberrations. Our main consideration in selecting this option was the relevance to possible clinical applications in the short term. The technological state of the art in ophthalmic optics and refractive surgery does not include the ability to correct for real-time wavefront errors in human eyes, although several methods already exist to permanently modify ocular wavefront aberrations. Such methods include custom-wavefront, wavefront optimized, aspheric and progressive optical designs in contact lenses, spectacle lenses, intraocular lenses, and corneal tissue ablation patterns using laser surgery. We thus had a strong interest in exploring the effects of static wavefront modifications generated by an adaptive optics system.
There was also a practical advantage in experimenting with static wavefronts. The application of dynamically adjusting wavefronts, using the procedure described in the Methods section, was likely to provide the most stable retinal images of the visual acuity test optotypes, but this procedure required illuminating the retina with an infrared beam for the wavefront sensor to monitor the eye's aberrations. In practice, although its center wavelength was theoretically outside of the visible range, this beam could always be seen by the subject as a red spot superimposed with the visual acuity test. The use of static wavefronts eliminated the need for illuminating the retina with infrared light and avoided disturbing the visual stimulation.
A difficulty in simulating static wavefronts is that the adaptive optics system must be accurately aligned with the pupil. According to Holladay et al,15 the correction of spherical aberration using aspheric intraocular lenses may become ineffective if the lenses are shifted off axis by more than 0.4 mm. De Brabander et al16 calculated that a custom-wavefront contact lens correction would be optically beneficial to keratoconic eyes if such lenses could be positioned with an accuracy of 0.5 mm or better. The tolerance to a lateral shift of the wavefront correction was larger in the case of normal eyes. In the current study, the positioning errors of the wavefront corrections did not exceed 0.3 mm. The precision in aligning our device was thus sufficient for higher order corrections to improve the optical performance of the eyes in this study.
Most adaptive optics systems have two limitations that prevent them from being used in clinical practice: (1) their relatively large size and (2) their limited range of correction. With electromagnetic adaptive optics technology, we have for the first time overcome these limitations by clinically using a compact adaptive optics instrument to accurately correct and generate aberrations of high magnitude. The range of correction provided by the electromagnetic adaptive optics technology was suitable for the simulation of customwavefront corrections in all of the eyes involved in our study. The deformable mirror was able to maintain the lower and higher order aberrations close to the desired levels during the visual acuity measurements.
However, one limitation of this system was that not all Zernike terms could be equally corrected and generated, as also discussed by Fern?ndez et al.4 The correction quality (eg, the residual RMS) of the internal adaptive optics system reached approximately 150 nm with the eyes that had very large aberrations. An imperfection in our experiment was that we calibrated the adaptive optics system using the largest available aperture (6.5-mm diameter) while we ran the closedloop using an artificial pupil diameter of 5 mm. Under these conditions, the behavior of the closed-loop algorithms was not optimal. The elimination of this problem by better-optimized software should lead to higher correction quality in cases with large aberrations.
For the measurement of visual acuity, we used the FrACT (Landolt C optotypes). Studies have shown that the FrACT presents the same results as conventional forced-choice chart testing, but in a shorter time.17 Reading and Weale18 demonstrated that the clinically used Snellen visual acuity test chart was comparable with the screen-based self-administered computerized Lando It-C test in 86 patients with ages ranging from 20 to 79 years. Ruamviboonsuk et al19 further demonstrated that the automated Landolt-C testing system was reproducible and comparable to the Early Treatment Diabetic Retinopathy Study (ETDRS) chart. These studies support the automated test as a reasonable alternative tool for measuring visual acuity outcomes.
In the current study, the impact of individually generated Zernike aberrations of various RMS magnitudes on the visual acuity of wavefront-corrected normal and highly aberrated eyes was evaluated using the crxl Adaptive Optics Simulator. We demonstrated that each individual mode of the Zernike expansion had a different impact on visual function as measured by a psychophysical test. The generation of individual Zernike terms at various magnitudes of RMS error from the fully corrected eye created a pattern of mean change in visual acuity that was uniquely suited to the Zernike term. With increasing RMS error, the logMAR acuity decreased disproportionately among terms, with the greatest change noted with spherical aberration and the least with trefoil. The term of greatest impact with relatively low RMS error (0.1 and 0.3 pm) was defocus, but with higher RMS error (0.9 ???), spherical aberration increased to meet defocus as the most influential term. This latter observation has great practical importance in clinical laser vision correction because of the high induction of spherical aberration with high myopic corrections.
The disproportionate visual impact of individual aberrations in our study concurs with the observation of Applegate et al,10,11 who reported that aberrations near the center of the pyramid-like Zernike table (eg, coma, spherical aberration, and secondary astigmatism) cause greater distortion of visual quality than those at the periphery of the table. Additionally, the correction of higher order monochromatic aberrations in white light has been performed using a deformable mirror by Yoon and Williams,7 estimating their differential impact on visual acuity and contrast sensitivity.
Considering these and other studies, it is clear that higher order ocular aberrations degrade the optical quality of the eye. In our study, we found a mean improvement of approximately one optotype chart line for all patients' eyes compared to their best sphero-cylinder spectacle correction after adaptive optics correction. Although the effect was statistically significant, it is probably not clinically significant for all patients. One might infer that some eyes will gain a greater visual benefit from a custom-wavefront LASIK procedure than others, and this, in part, will be due to the mode and magnitude of their individual aberrations.
The visual impact of higher order ocular aberrations of differing magnitude using an adaptive optics system could also help us to better understand and predict the visual outcomes of wavefront-guided procedures. Previous studies have reported that some combinations of Zernike modes are less deleterious than the individual effects of those aberrations by themselves.10,11'2021 Future studies using adaptive optics should explore the effects of combined Zernike aberrations to better simulate and predict the visual impact of laser vision correction and its optical complications and symptoms.
In our study, the two most highly aberrated eyes, which should have had the largest improvements, did not regain as many lines of visual acuity as expected. Possible reasons for this reduced correlation between the amount of corrected higher order ocular aberration and the improvement of visual function could be attributed to neuroadaption and also residual higher order ocular aberrations that remained uncorrected by the mirror. These differences between subjects might be related to the interpretation of the retinal image by the visual system. Because both optical and neural factors limited contrast sensitivity and visual resolution,2223 correction of the optical limitations could allow us to measure the neural component of vision or neural transfer function.
The evaluation of the effect of the accommodation during generation of individual Zernike modes may be another limitation of this study, especially in young patients. The examinations were reproducible and were performed in a dark room with a pupil diameter of at least 5 mm during the entire test.
Understanding the optical properties and aberrations of the eyes of different patients and then compensating for the eye's aberrations could enhance visual performance by greatly improving retinal image quality and optimizing surgical results. New metrics of the optical effects and the neural transfer functions that are indicative of visual function need to be developed. Future applications of a visual simulator include exploring visual function by means of measuring neural transfer function and screening candidates for custom- wavefront and presbyopic LASIK treatments.
The introduction of adaptive optics technology to clinical ophthalmic practice could provide a useful measurement tool for correcting aberrations and predicting the potential benefits of customized corrections.
1. Liang J, Williams DR, Miller DT. Supernormal vision and highresolution retinal imaging through adaptive optics. / Opt Soc Am A Opt Image Sci Vis. 1997;14:2884-2892.
2. Hofer H, Artal P, Singer B, Aragon JL, Williams DR. Dynamics of the eye's wave aberration. / Opt Soc Am A Opt Image Sci Vis. 2001;18:497-506.
3. Fernandez EJ. Ultrahigh resolution optical coherence tomography with adaptive optics: experimental results and limits to the performance. Presented at: Engineering the Eye II: Imaging the Retina; June 19-20, 2006; National University of Ireland, Galway, United Kingdom.
4. Fernandez EJ, Manzanera S, Piers P, Artal P. Adaptive optics visual simulator. J Refract Surg. 2002;18:S634-S638.
5. Makous W, Carroll J, Wolfing JI, Christie N, Williams DR. Retinal micro scotomas revealed with adaptive -optic s microflashes. Invest Ophthalmol Vis Sci. 2006;47:4160-4167.
6. Guirao A, Porter J, Williams DR, Cox IG. Calculated impact of higher-order monochromatic aberrations on retinal image quality in a population of human eyes. / Opt Soc Am A Opt Image Sci Vis. 2002;19:1-9.
7. Yoon GY, Williams DR. Visual performance after correcting the monochromatic and chromatic aberrations of the eye. / Opt Soc Am A Opt Image Sci Vis. 2002;19:266-275.
8. Applegate RA, Ballentine C, Gross H, Sarver EJ, Sarver CA. Visual acuity as a function of Zernike mode and level of root mean square error. Optom Vis Sci. 2003;80:97-105.
9. Applegate RA, Hilmantel G, Howland HC, Tu EY, Starck T, Zayac EJ. Corneal first surface optical aberrations and visual performance. J Refract Surg. 2000;16:507-514.
10. Applegate RA, Sarver EJ, Khemsara V. Are all aberrations equal? J Refract Surg. 2002;18:S556-S562.
11. Applegate RA, Marsack JD, Ramos R, Sarver EJ. Interaction between aberrations to improve or reduce visual performance. J Cataract Refract Surg. 2003;29:1487-1495.
12. Doshi JB, Sarver EJ, Applegate RA. Schematic eye models for simulation of patient visual performance. / Refract Surg. 2001;17:414-419. Erratum in: / Refract Surg. 2001;17:498-499.
13. Bach M. The Freiburg Visual Acuity test ? automatic measurement of visual acuity. Optom Vis Sci. 1996;73:49-53.
14. Bach M. Freiburg Visual Acuity, Contrast, and Vernier Test ('FrACT'). Available at: http://www.michaelbach.de/fract. Accessed 2007.
15. Holladay JT, Piers PA, Koranyi G, van der Mooren M, Norrby NE. A new intraocular lens design to reduce spherical aberration of Pseudophakie eyes. / Refract Surg. 2002;18:683-691.
16. de Brabander J, Chateau N, Marin G, Lopez-Gil N, Van Der Worp E, Benito A. Simulated optical performance of customwavefront soft contact lenses for keratoconus. Optom Vis Sci. 2003;80:637-643.
17. Mittelviefhaus K, Bach M, Jedynak A, Kommerell G. The Freiburg Vision Test. A computer-assisted procedure with sequential strategy [German]. Ophthalmologe. 1993;90:132-135.
18. Reading VM, Weale RA. Self-administered automatic sight-testing. A feasibility study. Doc Ophthalmol. 1993;83:43-54.
19. Ruamviboonsuk P, Tiensuwan M, Kunawut C, Masayaanon P. Repeatability of an automated Landolt C test, compared with the early treatment of diabetic retinopathy study (ETDRS) chart testing. Am J Ophthalmol. 2003;136:662-669.
20. Thibos LN, Hong X, Bradley A, Cheng X. Statistical variation of aberration structure and image quality in a normal population of healthy eyes. / Opt Soc Am A Opt Image Sci Vis. 2002;19:23292348.
21. McLellan JS, Prieto PM, Marcos S, Burns SA. Effects of interactions among wave aberrations on optical image quality. Vision Res. 2006;46:3009-3016.
22. Campbell FW, Green DG. Optical and retinal factors affecting visual resolution. / Physiol. 1965;181:576-593.
23. Charman WN, Chateau N. The prospects for super-acuity: limits to visual performance after correction of monochromatic ocular aberration. Ophthalmic Physiol Opt. 2003;23:479-493.
Means and Inter-individual Standard Deviations of the Measured Visual Acuity Data (logMAR)