Aberrations and visual performance following standard laser vision ...

Aberrations and Visual Performance FollowingStandard Laser Vision CorrectionSusana Marcos, PhDABSTRACTPURPOSE: To relate the change of ocular aberrationswith the change of visual performance producedby standard laser in situ keratomileusis(LASIK) for myopia.METHODS: Aberrations and visual performancewere measured before and after surgery in 22 eyes.Total aberrations were measured using a laser raytracing technique. Corneal aberrations wereobtained using a commercial videokeratoscope andcustom software. Visual performance was evaluatedin terms of best spectacle-corrected contrastsensitivity and high-contrast visual acuity.RESULTS: The amount of total and corneal aberrationsincreased with LASIK in all except twoeyes. In general, the total and corneal aberrationpatterns were well-correlated after LASIK.However, the anterior corneal surface alone did notcapture all the information (ie, possible changesinduced on the posterior corneal surface or interactionsbetween the different ocular components),which suggests that total aberrations are betterpredictors of visual performance. The decrease ofthe modulation transfer function (both in terms ofarea under the curve or as a function of spatial frequency)with LASIK accounted for most of thedecrease in contrast sensitivity. High-resolutionvisual acuity was not very sensitive to the LASIKinducedchanges in image quality.From the Instituto de Optica “Daza de Valdés,” Consejo Superior deInvestigaciones Científicas, Madrid, Spain.The author has no proprietary interest in the methods presentedherein.Presented at the 2nd International Congress of Wavefront Sensing andAberration-free Refractive Correction, in Monterey, CA, on February9-10, 2001.The author acknowledges Dr. Esther Moreno-Baniuso for herinestimable contributions to the total aberration measurements; Mr. SergioBarbero for invaluable work in generating custom software for topographicanalysis; Ms. Lourdes Llorente for data analysis; Ms. GuadalupeRodríguez and Mr. Raúl Martín for clinical data collection, and collaborativerefractive surgeon, Dr. Jesús Merayo-Lloves, from the Instituto deOftalmobiología Aplicada, Universidad de Valladolid; Carl-Zeiss Spainfor the loan of Atlas Humphrey Corneal Topography System and partialfunding of a research fellowship; and CAM08.7/0010.1/2000 andConvenio de Cooperación USA-Spain for research funding.Correspondence: Susana Marcos, PhD, Instituto de Optica, ConsejoSuperior de Investigaciones Científicas (CSIC), Serrano 121, 28006Madrid, Spain. Tel: 34.915616800 x2306; Fax: 34.915645557; E-mail:susana@io.cfmac.csic.esCONCLUSION: Wavefront and total aberrationmeasurements, and in particular a combination ofthe two techniques, provide useful information forunderstanding the optical changes induced by standardrefractive surgery. Predictions of the changein modulation transfer function derived from waveaberrations matched measurements of change incontrast sensitivity. [J Refract Surg 2001;17:S596-S601]The development of videokeratoscopes capableof recording corneal shape in detail, and aberrometersthat measure the wave aberration ofthe optics of the entire eye, have revealed thatalthough standard laser refractive surgery eliminatesconventional refractive errors, higher ordererrors (particularly spherical aberration) are typicallyinduced. 1-3 Numerous reports in the literatureevaluate visual performance after laser in situ keratomileusis(LASIK) 4,5 , but only a few attempt tocharacterize the correlation between visual performanceand optical aberrations (in particular, theaberrations of the anterior corneal surface) inducedby refractive surgery. 6 We compared corneal andtotal aberrations as well as visual performancebefore and after standard LASIK for myopia in agroup of patients. Although most of the changesinduced by LASIK occur on the anterior corneal surface,the total wave aberration captures possiblechanges of the posterior corneal surface and interactionsbetween the different optical components,and therefore should predict visual outcomes moreaccurately. We report a good correlation between thedecrease of contrast sensitivity function afterLASIK and the decrease in the modulation transferfunction (computed from the total wave aberration).PATIENTS AND METHODSTotal Aberrations: Laser Ray TracingWave aberrations were measured using laser raytracing, a technique developed by Navarro and colleaguesat the Instituto de Optica, Madrid, Spain. 7S596 Journal of Refractive Surgery Volume 17 September/October 2001

Aberrations and Visual Performance Following Standard Laser Vision Correction/MarcosFigure 1. Total and corneal wave aberrationmaps, before and after standard LASIK formyopia for A) eye #14, B) eye #16, C) eye#18, and D) eye #19. Total aberrations measuredwith laser ray tracing. Corneal aberrationmeasured using a commercial cornealvideokeratoscope and custom software. Tilts,defocus, and astigmatism were cancelled.Pupil diameter = 6.5 mm.This technique provides similar results to Shack-Hartmann and the Spatially Resolved Refractometerin normal eyes, and is appropriate in highlyaberrated eyes. 8 In this technique, a set of parallelrays is projected sequentially through 37 differentpupil positions (forming a hexagonal array thatsamples a 6.5-mm pupil). The corresponding aerialretinal images are collected on a high-resolutionCCD camera. The deviation of each centroid fromthe principal ray is proportional to the local slope ofthe wave aberration. The wave aberration isobtained from the derivatives using a least-meansquaresprocedure and described as a seventh orderZernike polynomial expansion. Measurements wereperformed on dilated pupils. The patient fixatedfoveally and the eye’s pupil was aligned to the opticalaxis of the instrument. Figure 1 (A, B, C, and D,left panels) shows examples of total wave aberrationsfor four patients before and after LASIK. Rootmean-square(RMS) wavefront error is typicallyused as an optical quality metric.Corneal Aberrations From Corneal TopographyCorneal aberrations were estimated using aplacido-disk-based videokeratoscope (Atlas MastervueCorneal Topography System, HumphreyInstruments-Zeiss, San Leandro, CA) and customsoftware developed by Sergio Barbero 9 (Instituto deOptica, CSIC, Madrid, Spain) in Matlab(Mathworks, Nathick, MA) and Zemax (FocusSoftware, Tucson, AZ). In brief, the slopes of thecorneal wave aberration were calculated by virtualray tracing on the corneal elevation, and cornealaberrations were then obtained using a similar procedureto that described for the total aberrations.Since corneal aberrations use the corneal reflex as areference and total aberrations are measured withrespect to the pupil center, a realignment algorithmwas applied to allow a direct comparison of cornealand total aberration maps. The small tilt betweenthe videokeratographic axis and the line of sightwas not considered. A series of corneal aberrationmaps were obtained over 6.5 mm, shifting the centerat 0.1-mm steps. The difference, total – cornealRMS, was then computed as a function of pupil location.These maps are smooth and show a welldefined minimum, slightly decentered from thecorneal reflex, which was used as a common axis.Figure 1 (A, B, C and D, right panels) shows examplesof corneal wave aberrations for four patientsbefore and after LASIK.Journal of Refractive Surgery Volume 17 September/October 2001 S597

Aberrations and Visual Performance Following Standard Laser Vision Correction/MarcosVisual Performance: Contrast Sensitivity Function andVisual AcuityVisual performance was assessed by means ofcontrast sensitivity function (with best spectaclecorrection) and best spectacle-corrected high-contrastvisual acuity. Contrast sensitivity was measuredusing a standard CVS-1000E chart (VectorVision, Arcanum, OH). This chart uses vertical sinusoidalgrids (at 3, 6, 12, and 18 c/deg), a forceddouble-alternative choice paradigm, and a calibratedluminance of 85 cd/m 2 . Figure 2 shows the averagecontrast sensitivity function (22 patients) beforeand after LASIK. High contrast visual acuity wasmeasured using a conventional Snellen chart.Figure 2. Contrast sensitivity function (average of 22 eyes) beforeand after LASIK, measured with the CSV-1000E chart. Lines arethird order polynomial fits to the data, error bars represent standarderror of the mean, with best spectacle-correction and undilutedpupil.Patients and ProceduresTwenty-two eyes from 12 patients (mean age,28 ± 5 yr; preoperative spherical error, -2.50 to-13.00 diopters [D]) participated in the study.LASIK was performed using a narrow-beam, flyingspotexcimer laser (Chiron Technolas 217-C; Bausch& Lomb, Surgical, Dornach, Germany), assisted byan eye-tracker. The flap diameter (performed with aHansatome microkeratome) was 8.5-mm and theintended depth 180 µm. Photoablation was appliedto a 6-mm optical zone, with a transition zone of 9mm. The LASIK procedures were conducted at theInstituto de Oftalmobiología Aplicada, Universidadde Valladolid, Spain, by Dr. Jesús Merayo-Lloves.Total aberrations were measured about 1 monthbefore and between 1 and 3 months after LASIK atthe Instituto de Optica, Madrid, Spain. Data weretypically the average of five sets of measurements.Corneal topography was performed during the sameexamination session. Visual performance measurementswere conducted at the Instituto deOftalmobiología Aplicada, Universidad deValladolid, by Raúl Martín and GuadalupeRodríguez, before and between 6 months and 1 yearafter LASIK.Figure 3. Root-mean-square (RMS) wavefronterror before (gray bars) and after (black bars)LASIK, for total third and higher order aberrations.Pupil diameter = 6.5 mm. Eyes are sorted byincreasing preoperative spherical error. Adaptedfrom Moreno-Barriuso et al. 3S598 Journal of Refractive Surgery Volume 17 September/October 2001

Aberrations and Visual Performance Following Standard Laser Vision Correction/MarcosFigure 4. Increment of total aberrations (third order and higher) withLASIK versus increment of corneal aberrations. The dotted line indicatesa linear function through the origin (y=x). Most data points areabove this line, indicating a higher increase of corneal aberrations.Coefficient of correlation r=0.73, P=.0024.RESULTSChange of Total Aberrations With LASIKAs has been reported, total aberrations increasefollowing standard LASIK for myopia. 3 Figure 3shows the root-mean-square (RMS) wavefront errorbefore and after LASIK for third order and higheraberrations (ie, excluding tilt, defocus, and astigmatism),for a 6.5-mm pupil. Eyes are sorted byincreasing preoperative refraction. The averageincrease was 1.9 times, and the effect was more pronouncedfor the highest preoperative myopes.Spherical aberration increased by a factor of 3.9. 3 Asshown in Figure 1, this was a common trend in mosteyes (central area of positive aberrations surroundedby a ring of negative aberration). Third orderterms (including coma) 1 increased by a factor of 2,and were likely associated with a decentration of theablation pattern (the central area appeared decenteredin some cases, as shown in Figure 1). 3Aberrations increased significantly in all eyesexcept for two (eyes #10 and #11).Do Corneal Aberrations Change Similiar to TotalAberrations After LASIK?Figure 1 shows that although corneal and totalaberrations are in general different in normal eyes(prior to surgery), they show a high degree of similarityafter surgery. 10 Although there is a good correlationbetween corneal and total aberrations aftersurgery (r=0.97, P < .0001), a direct comparisondemonstrates the following 10 : 1) The spherical aberrationinduced in the anterior surface of the corneasignificantly exceeds that induced in the whole eye.This attenuation is likely produced by a sphericalaberration of negative sign induced on the posteriorsurface of the cornea (related to the reported forwardshift of the posterior corneal surface followingrefractive surgery). 2) The correlation of the incrementof total aberration with the increment ofcorneal aberrations (Fig 4) is lower (r=0.73, P =.0024) than the correlation of postoperative totaland postoperative corneal aberrations. This suggestsa significant role of the interaction of cornealand internal aberrations prior to surgery, and insome cases relevant after surgery. For example, therelative amount of corneal and internal (probablycrystalline lens) aberrations prior to surgeryexplains the surprisingly good outcome encounteredin eyes #10 and #11. 10 In general, total aberrationsshould be better related to visual performance thanaberrations of the anterior corneal surface alone,since they take into account possible changes of theposterior corneal surface and interaction with otherocular components (crystalline lens and pupil).Predictions From Aberrometry and PsychophysicalMeasurements of Visual PerformanceContrast sensitivity function represents the contrastdegradation imposed by the optics and posteriorvisual processing as a function of spatial frequency.Since only the optics are modified with LASIK,one expects any change induced in contrast sensitivityfunction to be due to a change only in the opticalsystem, and more specifically in the modulationtransfer function (MTF) of the eye. The Strehl ratio(normalized volume under the MTF) is an alternativeglobal image quality parameter to the rootmean square error. Both metrics are, in general,well-correlated. The MTF can be obtained easilyfrom the wave aberration using Fourier optics. Forbetter comparison with the one-dimensional contrastsensitivity function (for vertical gratings), weused the horizontal section of the MTF. We computedthe MTF for a 3-mm diameter pupil, to simulatea closer condition to the contrast sensitivity functionmeasurement, performed with an undilated pupil.The area under these curves was computed, usinglinear units for both spatial frequency and contrastunits, a linear interpolation, and integratingJournal of Refractive Surgery Volume 17 September/October 2001 S599

Aberrations and Visual Performance Following Standard Laser Vision Correction/MarcosABCVA = 1BCVA = 0.9BCVA = 1BFigure 5. A) Modulation transferratio postoperative/preoperative(solid line) and contrast sensitivityratio postoperative/preoperative(circles and dashed line) as afunction of spatial frequency. Dataare average of 22 eyes. Pupildiameter = 3 mm for the modulationtransfer function (MTF), andundilated for the contrast sensitivityfunction (CSF). B) Simulation ofretinal image quality (obtained byconvolution of the 20/20 line on aSnellen chart and the correspondingpoint spread function) and clinicalmeasurements of best spectacle-correctedvisual acuity, forthree eyes (#14, #21, #11) withdifferent surgical outcomes.between 3 and 18 c/deg. The area under the MTF(average of 22 eyes) decreased by a factor of 1.38after LASIK, and the area under the contrastsensitivity function (average of 22 eyes) decreasedby a factor of 1.51 after LASIK. This indicates thatthe average image contrast degradation estimatedfrom wave aberration data accounts for most of thedecrease in contrast sensitivity in this spatial frequencyrange. Figure 5 shows the average contrastratio before and after surgery for both modulationtransfer and contrast sensitivity as a function ofspatial frequency. Again, both functions tend todecrease similarly with LASIK. The fact that contrastsensitivity function seems to suffer a slightlylarger degradation could be due to the fact thatpupils were larger than 3 mm during the contrastsensitivity function measurements, that the visualperformance and aberration measurements werenot collected on the same day, that MTF computationswere based on monochromatic aberrationswhereas contrast sensitivity function was measuredin polychromatic light, or that other factors apartfrom the aberrations (such as haze) may also contributeto contrast degradation. The eye in whichthe root mean square decreased (although not significantly)after LASIK experienced a slightincrease in the MTF postoperative/preoperativeratio, as well as a contrast sensitivity improvementat certain spatial frequencies (3 and 12 c/deg).We also compared visual performance (best spectacle-correctedvisual acuity [BSCVA]) with simulationsof retinal images of a Snellen chart. Thesewere generated by convolution of point-spread-functions(computed from the wave aberrations) withimages of the Snellen chart. These simulations donot represent vision of patients, but provide an ideaof the image quality on the retinal plane. Figure 5Bshows simulations of the 20/20 line of a Snellenchart for three eyes (#14, #21, #11) after LASIK.Clinical BSCVA measurements for these eyes arereported. In many cases, such as eye #14, a cleardegradation of retinal image quality is not associatedwith a line loss, indicating that unlike contrastsensitivity, high-contrast visual acuity is not a verysensitive measurement to evaluate the changesinduced by refractive surgery.From measurements of total and corneal aberrations,contrast sensitivity and visual acuity measuredin a group of eyes before and after surgery, weconclude that:1) Both total and corneal aberrations increasesignificantly following standard myopic LASIK.2) Total aberrations should be better correlated tovisual performance than corneal aberrations alone.Although most of the changes occur on the anteriorcorneal surface, in general corneal aberrations arenot sufficient to understand the changes induced byrefractive surgery. Preoperative interaction of thecorneal and internal optics, and possible changesinduced on the posterior corneal surface also contributeto the total aberration pattern.3) Most of the decrease in contrast sensitivity canbe accounted for by the decrease in the modulationtransfer function (computed from the total waveaberration)4) High contrast visual acuity is not a very sensitivemeasurement of image quality.REFERENCES1. Oshika T, Klyce SD, Applegate RA, Howland HC, ElDanasoury MA. Comparison of corneal wavefront aberrationsafter photorefractive keratectomy and laser in situkeratomileusis. Am J Ophthalmol 1999;127:1-7.S600 Journal of Refractive Surgery Volume 17 September/October 2001

Aberrations and Visual Performance Following Standard Laser Vision Correction/Marcos2. Mierdel P, Kaemmerer M, Krinke HE, Seiler T. Effects ofphotorefractive keratectomy and cataract surgery on ocularoptical errors of higher order. Graefes Arch Clin ExpOphthalmol 1999;237:725-729.3. Moreno-Barriuso E, Merayo-Lloves J, Marcos S, Navarro R,Llorente L, Barbero S. Ocular aberrations before and aftermyopic corneal refractive surgery: LASIK-induced changesmeasured with Laser Ray Tracing. Invest Ophthalmol VisSci 2001;42:1396-1403.4. Holladay JT, Dudeja DR, Chang J. Functional vision andcorneal changes after laser in situ keratomileusis determinedby contrast sensitivity, glare testing and cornealtopography. J Cataract Refract Surg 1999;25:663-669.5. Mutyala S, McDonald M, Scheinblum K, Ostrick M, Brint S,Thompson H. Contrast sensitivity evaluation after laser insitu keratomileusis. Ophthalmology 2000;107:1864-1867.6. Applegate RA, Howland HC, Sharp RP, Cottingham AJ, YeeRW. Corneal aberrations and visual performance after radialkeratotomy. J Refract Surg 1998;14:397-407.7. Navarro R, Losada MA. Aberrations and relative efficiencyof light pencils in the living human eye. Optom Vis Sci1997;74:540-547.8. Moreno-Barriuso E, Marcos S, Navarro R, Bums SA.Comparing Laser Ray Tracing, Spatially ResolvedRefractometer, and Hartmann-Shack sensor to measure theocular wavefront aberration. Optom Vis Sci 2001;78:152-156.9. Barbero S, Marcos S, Martin R, Llorente L, Moreno-Barriuso E, Merayo-Lloves JM. Validating the calculation ofcorneal aberrations from corneal topography: a test on keratoconusand aphakic eyes. Invest Opthalmol Vis Sci2001;42(suppl):S894.10. Marcos S, Barbero S, Llorente L, Merayo-Lloves J. Opticalresponse to myopic LASIK surgery from total and cornealaberration measurements. Invest Ophthalmol Vis Sci,in press.Journal of Refractive Surgery Volume 17 September/October 2001 S601

Running title/Author602 Journal of Refractive Surgery Volume ?? Month/Month 1997