Secondary Logo

Journal Logo


Color discrimination by patients with different types of light-filtering intraocular lenses

Ao, Mingxin MD; Chen, Xiaoyong MD; Huang, Chen MD; Li, Xuemin MD; Hou, Zhiqiang MD; Chen, Xue MD; Zhang, Chun MD; Wang, Wei MD

Author Information
Journal of Cataract & Refractive Surgery: March 2010 - Volume 36 - Issue 3 - p 389-395
doi: 10.1016/j.jcrs.2009.09.038
  • Free


Optical radiation includes ultraviolet (UV) (200 to 400 nm), which does not contribute to vision and damages the retina with very intense exposure, and visible light (400 to 760 nm).1 The cornea can block UV radiation with wavelengths shorter than 300 nm and the crystalline lens with wavelengths longer than 300 nm.2 Despite the natural filtering ability of the cornea, the photoreceptors and retinal pigment epithelial (RPE) cells are still exposed to high concentrations of light and oxygen, the main factors in photic retinopathy.3,4 With aging, the intraretinal defenses against photic retinopathy decline, making the retina more vulnerable.3,5 Meanwhile, ocular media defenses in the aging crystalline lens increasingly block more potentially phototoxic visible short-wavelength light, the so-called blue light. This light can injure the retina; UV and blue–green light can cause retinal phototoxicity and may play a role in the progression of age-related macular degeneration (AMD). The severity of UV blue-light phototoxicity increases with decreasing wavelength, and its action spectrum covers ultraviolet light (wavelengths <400 nm), violet light (wavelength 400 to 440 nm), and blue light (wavelength 440 to 500 nm). On the contrary, the action spectrum of the blue–green type of phototoxicity peaks in the range of blue light, green light, and yellow light.4,6–9

Cataract surgery with intraocular lens (IOL) implantation may disturb this natural balance because the IOL transmits more optical radiation to the retina than the aging crystalline lenses, which can compromise the ocular media's defenses against photic retinopathy.10 To imitate the UV radiation–blocking property of the natural lens, IOLs with UV radiation–filtering chromophores have been routinely used in cataract surgery since the early 1980s.11,12 In an attempt to simulate the protective properties of a natural lens, IOLs filtering visible short-wavelength blue light (from 400 to 500 nm) were introduced in the 1990s.13,14 These types of IOLs could have a cytoprotective effect against light-induced stress on RPE cells, and this protection might be more essential in the elderly population, especially in eyes with early signs of AMD.1,15

The controversy about blue light–filtering IOLs stems from the fact that the filtered visible blue light, especially that with a wavelength from 440 to 500 nm, plays an important role in the physiology of scotopic, contrast, and color vision and in circadian photoentrainment.9,16 Several studies have evaluated ways to balance photoprotection and photoreception by replacing the aging opacified crystalline lens with an IOL without compromising the protection afforded by the crystalline lens.9,17,18 The controversy over visual function in eyes with blue light–filtering IOLs mainly centers on performance under scotopic and mesopic conditions. Therefore, there is increasing interest in photochromic blue light–filtering IOLs that transmit nearly all visible light in dim environments for optimum visual performance and filter moderate amounts of high-energy short-wavelength visible light in bright environments to protect against phototoxicity.17

This study compared visual quality with a photochromic blue light–filtering IOL and with a conventional yellow-tinted blue light–filtering IOL. An IOL that filters UV light only was used as a control. Because the effect of conventional blue light–filtering IOLs on color perception is controversial19,20 and there is a reported case of extraction of a conventional yellow-tinted IOL because of an intolerable imbalance in color perception between the eye with that IOL and the fellow eye with a UV filtering–only IOL,21 our study focused on color perception under photopic conditions and mesopic conditions.


This study evaluated patients with uncomplicated senile cataract who had phacoemulsification with IOL implantation. An institutional review board approved study protocol, and all patients provided informed consent.

Inclusion criteria included patient satisfaction with overall surgery, a corrected distance visual acuity (CDVA) of 20/25 or better, no history of color vision deficiency, and passing color-screening test results (Ishihara plate test) 3 months postoperatively. Exclusion criteria were ocular disease other than senile cataract, history of ocular surgery, posterior capsule opacification, and previous or current use of medications known to cause color-vision deficiencies.

Visual functional evaluations were performed 3 months after surgery. The examinations included were open to all postoperative cataract patients; only patients meeting the inclusion criteria and willing to complete the specific tests were enrolled in the study.

Intraocular Lenses

Patients received 1 of 3 types of IOLs. Transmission characters of these 3 types of IOLs are shown in Figure 1.

Figure 1
Figure 1:
Light-transmission curve of the 3 IOLs. Blue line: Control IOL. Red line: Photochromic blue light–filtering IOL fully activated by UV radiation. Yellow line Yellow-tinted UV- and blue light–filtering IOL.

The photochromic group received a photochromic blue light–filtering IOL (MatrixAcrylic Aurium 400 IOL, Medennium). This 3-piece foldable IOL filters short-wavelength light (400 to 550 nm) under UV light conditions and acts as a conventional UV-filtering IOL when there is no UV light. According to the manufacturer, the IOL can activate under radiation with a wavelength of 429 nm and fully activate or fade within 200 seconds (Figure 2). In the preliminary research designed to evaluate the IOL's feasibility and biocompatibility, the wavelength of the projected light was 365 nm.22 The IOL has an overall diameter of 12.5 mm, an optic diameter of 6.0 mm, and a refractive index of 1.56.

Figure 2
Figure 2:
Activation kinetics and fade kinetics of the photochromic blue light–filtering IOL.

The yellow-tinted group received a UV-filtering and blue light–filtering IOL (AcrySof Natural SN60AT IOL, Alcon, Inc.). The 1-piece foldable IOL contains a covalently bound light-yellow chromophore that filters UV light from 300 to 400 nm and short-wavelength visible light from 400 to 500 nm. It has an overall diameter of 13.0 mm, an optic diameter of 6.0 mm, and a refractive index of 1.55.

The control group received a conventional IOL that filters UV light only (MatrixAcrylic 401 IOL, Medennium). The physical characteristics of the IOL are identical to those of the MatrixAcrylic Aurium 400 IOL except it does not have the photochromic blue light–filtering property.

Surgical Technique

The same surgeon (W.W.) performed all cataract extractions using topical anesthesia and a 3.2 mm clear corneal temporal incision. Phacoemulsification was followed by irrigation and aspiration of cortex and IOL implantation in the capsular bag.

Visual Acuity

Corrected distance visual acuity was measured using a Snellen chart under standard conditions. Monocular Snellen visual acuity values were transformed to logMAR values for analysis.

Monocular visual acuity under mesopic conditions was performed with the best spectacle correction (obtained from refractive examination) using a Snellen chart with 40 lux illumination. Before the visual acuity test, patients were allowed to rest and adapt to the light level for at least 10 minutes.23 Monocular Snellen visual acuity values were transformed to logMAR values for analysis.

The same technician performed the refractive examinations. The refractive technician was masked to IOL type.

Color Perception

Color perception was tested using the Farnsworth-Munsell 100-hue (FM 100-hue) test, a widely used tool to detect chromatic discrimination abnormalities and provide noninvasive assessment of eye disorders related to color perception. The test contains removable color reference caps. Color vision abnormalities and aptitude are detected by the patient's ability to place the color caps in order of hue.

Patients rested for 15 minutes before taking the FM 100-hue test.24 The test was performed binocularly in patients who had bilateral cataract surgery, with spectacle correction if necessary. In patients with monocular surgery, the test was performed in the eye with the IOL, with spectacle correction if necessary and with the phakic eye occluded. The test was performed according to the manufacturer's instruction. When the test was completed, the examiner recorded the results (order of color caps) without disclosing them to the patient.

Patients took the entire test under mesopic conditions and photopic conditions with an interval of 10 minutes between the 2 parts. The photopic portion was performed with illumination of approximately 1000 lux (normal room lights, room door shut, window curtains opened); illumination was provided by a whole-spectrum light. The mesopic portion was performed with an illumination level of approximately 40 lux (room lights turned off, room door shut, window curtains closed); the axle of the whole-spectrum light was turned to prevent it from projecting directly onto the color caps.

The error scores were calculated using FM 100-hue test scoring software (version 3.0) and the standard method. The total error score and the score of every color cap were recorded with the corresponding illumination condition. To detect color perception around the hue circuit of the FM 100-hue, partial error scores in the 10 color bands were calculated.

Patient-Reported Visual Outcomes

Subjective patient visual outcomes were assessed 3 months after surgery using the National Eye Institute Visual Functioning Questionnaire (NEI VFQ). The questionnaire measures multiple domains of vision-related functioning and health-related quality of life and has been validated for use in patients with a variety of eye conditions, including cataract.25 It has also been used to assess patient-reported visual outcomes in patients with different types of IOLs.26 This study used version 2000 of the 25-item NEI-VFQ (NEI VFQ-25). The examiner read and explained each item of the questionnaire to patients using a uniform script. The mean score and subheading scores were calculated according to the NEI VFQ-25 scoring algorithm.

The same resident administered the visual acuity test, questionnaire, and FM 100-hue test. The resident was masked to IOL type.

Statistical Analysis

Statistical analysis was performed using SPSS for Windows software (version 11.5, SPSS, Inc.). The mean, standard deviation, and median values were calculated. The Kolmogorov-Smirnov test was used to test the normality of data. The chi-square test was used to test the homogeneity of variance of 1-way analysis of variance (ANOVA). The 1-way ANOVA was used for data with normal distribution and chi-square test results. The Kruskal-Wallis test (analysis within groups) and Mann-Whitney U test (analysis between groups) was used to analyze other scale values. The Pearson chi-square test was adopted to analyze proportions of nominal values. The level of statistical significance was P<.05; for between-group comparisons, the value was corrected to P<.017 (0.05 of 3) to maintain a whole probability of a type 1 error of less than 0.05.


Patient Demographics

The study enrolled 43 patients. There were 15 patients in the photochromic group, 13 in the yellow-tinted group, 15 in the control group. Table 1 shows the patients' demographics by IOL group. There were no statistically significant differences between the groups in patient demographics or visual acuity. There were no significant intraoperative or postoperative complications.

Table 1
Table 1:
Demographic characteristics of patients by IOL group.

Color Perception

There were no statistically significant differences between groups in the total error scores on the FM 100-hue test under photopic conditions (P = .353) or mesopic conditions (P = .082). Partial error scores in the green to blue–green color band and blue–green to blue color band under mesopic conditions showed statistically significant differences within each group (P = .005 and P = .030, respectively). The perception in the green to blue–green color band was statistically significantly different between the photochromic group and the control group (P = .002 and P = .015) but not between the photochromic group and the control group (P = .678). In the blue–green to blue color band, between-group differences were marginally significant between the control group and the yellow-tinted group because of the corrected P value (P = .017). There were no statistically significant differences between the 3 groups in partial error scores in other color bands under mesopic conditions or in all partial error scores under photopic conditions. Table 2 shows the partial error scores in the 10 color bands under photopic conditions and Table 3, under mesopic conditions.

Table 2
Table 2:
Error scores in color bands on the FM 100-hue test under photopic conditions.
Table 3
Table 3:
Error scores in color bands on the FM 100-hue test under mesopic conditions.

Patient-Reported Visual Outcomes

The score for 4 subheadings (social function, dependency, color perception, and peripheral vision) on the NEI VFQ-25 was 100 in all cases. The subheading that describes quality of life related to driving was excluded because the patients in the study had no experience driving. The mean NEI VFQ-25 scores were comparable between the 3 IOL groups (P = .774). There were no statistically significant differences between the groups in subheading scores (Table 4).

Table 4
Table 4:
Subheading scores on the NEI VFQ-25.


In the elderly population with normal visual acuity, there is a tendency with age toward increasing error scores on the blue–yellow axis and red–green axis of the FM 100-hue test, indicating more errors in blue–yellow perception with aging.27 Changes in color perception may be more sensitive when there are subtle changes in the ocular system, such as changes in the transmitting property of ocular media.

Under photopic conditions, overall color perception results, described by the total error scores of the FM 100-hue test, showed no significant differences between the photochromic group and the yellow-tinted group and control groups. This result agrees with findings of Rodríguez-Galietero et al.,24 who used the FM 100-hue test to evaluate photopic color discrimination in eyes with the conventional blue light–filtering IOL we used in our study and an IOL with UV filtering only. Bhattacharjee et al.28 studied the effects of the conventional blue light–filtering IOL on photopic color vision and concluded that the IOL provides color perception equivalent to that of a UV-filtering only IOL.

In our study, the within-group total error scores on the FM 100-hue test under mesopic conditions were similar. There was no significant difference in discrimination of the 10 bands of the hue circuit between the photochromic group and the control group; the results in the green to blue–green band in both groups were statistically significantly different from those in the yellow-tinted group. In the blue–green to blue band, a band next to the green to blue–green band, the control group and yellow-tinted group had marginally different performance, although there was no difference between the latter group and the photochromic group.

Several studies have assessed the effects of blue light–filtering IOLs on overall and specific color perception under mesopic condition. Using the FM 100-hue test, Cionni and Tsai19 examined color perception in patients with bilateral implantation of an IOL that filters both UV and blue light or an IOL that filters UV light only. The overall color perception, perception on the red–green axis, and perception on the blue–yellow axis under photopic conditions and mesopic conditions were similar between the 2 groups. In patients with a yellow-tinted blue light–filtering IOL, Mester et al.20 found significantly reduced color perception for blue under mesopic conditions over a 12-month follow-up and under photopic conditions over 6 months; however, all measured total error scores were within the normal range and no patient reported impaired color vision. In their study, Mester et al. paid specific attention to performance on box 3 and box 4 of the FM 100-hue test and evaluated blue light–related color perception by separately analyzing partial error scores in these 2 boxes. In a study by Wirtitsch et al.,29 a higher foveal threshold was detected in patients with blue light–filtering IOLs using short-wave automated perimetry.

In the present study, color discrimination was studied more intensively by analyzing performance on the 10 color bands of the hue circuit in FM 100-hue. This led to analysis of subtle perception against the background of comparable overall color discrimination.

According to theories of color perception, 3 types of cones mediate color perception. Short-wavelength cones are most sensitive to wavelengths around 420 nm. Middle-wavelength cones are most sensitive to wavelengths around 530 nm. Long-wavelength cones are most sensitive to wavelengths around 560 nm. Filtering light with a wavelength from 400 to 500 nm would affect the function of short-wavelength cones the most and that of middle-wavelength cones to an extent. Because of the Purkinje shift, the characteristic changes in sensitivity of the eye when transitioning from photopic vision to scotopic vision, which are caused by filtered blue light, become more obvious under mesopic conditions.

In the Munsell color system, from which the FM 100-hue test color samples were selected, the green to blue–green band corresponds to colors from green to green–blue and the blue–green to blue band corresponds to colors from green–blue to blue. Here, the green–blue color originates from green combined with a similar amount of blue. To evaluate the relationship between discrimination of these 2 color bands and the light transmitted by the IOLs, another color system that describes the relationship between different color bands and light of different wavelengths must be used. One possibility is the Commission Internationale de l'éclairage (CIE) system. In the CIE chromaticity diagram, the green to blue–green band and blue–green to blue band correspond to 3 color regions: the blue–green region with a main wavelength of 495 nm, the blue–green region with a main wavelength of 490 nm, and the green–blue region with a main wavelength of 485 nm. Therefore, the differences between the 3 IOL groups in the green to blue–green band and blue–green to blue color band suggest that color perception in pseudophakic eyes is modified by filtering light with a wavelength in the range of 480 to 500 nm. Light with a wavelength from 480 to 500 nm might be essential for subtle color vision in the color bands from green to blue on the FM 100-hue test under mesopic conditions. Because of its actinic property, the photochromic IOL had transmission similar to that in the control IOL group (UV filtering only) under mesopic conditions with little UV radiation. Thus, the photochromic group and control group had different results than the yellow-tinted IOL group.

In the present study, discrimination of color bands with shorter wavelengths (violet light) was comparable between the 3 groups. This result is consistent with the theory that violet light (wavelength 400 to 440 nm) contributes little to useful vision and that it might span the wavelengths in the visible spectrum that are most dangerous to the retina.30 Many studies have focused on the action spectrum of phototoxicity to the retina. A study by van de Kraats and van Norren18 found a reduction in blue-light damage in the signal from shortwave-sensitive cones that was caused by several commercially available IOLs and the natural eye media (aged 20 and 70 years). The authors suggest that a sharp cutoff filtering near 445 nm provides better performance by optimizing the balance between light reception and light protection. By computing pseudophakic action spectra and retinal phototoxicity, Carson et al.31 concluded that a filtering property in the range of 420 to 480 nm is the most useful criterion for gauging protection by IOLs. From the action spectra of lipofuscin adjusted for the transmission characteristics of ocular media for people in their 80s, Margrain et al.1 found 450 nm to be the peak of the action spectra for lipofuscin-mediated blue-light damage. These findings, combined with the present results, indicate that using a more specific filtering spectrum to avoid modifying color discrimination would not conflict with the concept of photoprotection.

To determine whether subtle differences in color perception cause subjective discomfort, we evaluated patient-reported visual outcomes using the NEI VFQ-25. The social function and dependency results indicated that all patients obtained an optimum overall visual outcome. The full score for the subheading color vision showed that the subtle differences in mesopic color perception led to no subjective perceptible disturbances. These agree with the results reported by Espindle et al.,32 who found that a blue light–filtering IOL improved visual-related and health-related quality of life in a manner similar to that of an IOL that filtered blue light only.

In conclusion, filtering short-wavelength (480 to 500 nm) visible light under mesopic conditions could modify color perception in the green-to-blue bands. Because of its actinic blue light–filtering property, the photochromic IOL might avoid this kind of color modification. However, based on our results, the subtle differences in color perception neither disturb objective overall color discrimination nor cause subjective discomfort. Similar to IOLs that filter UV light only, photochromic IOLs and yellow-tinted blue light–filtering IOL could improve patient-reported visual outcomes in cases of uncomplicated senile cataract. Further study should be performed to evaluate the retinal reaction to changes in light transmitted by different IOLs and the long-term retinal-protecting property of photochromic blue light–filtering IOLs. Furthermore, when evaluating visual effects related to IOL-filtering properties, proper lighting in the examination environment is key. In the transition from the photopic condition to the mesopic condition, there are changes in the composition of light projected onto the retina through the ocular media, including the IOL. Therefore, future studies might examine the effects of changes in IOL-filtering properties on retinal signal reception in the transition from photopic to mesopic conditions by simulating the optical pathway through the ocular media to the retina.


1. Margrain TH, Boulton M, Marshall J, Sliney DH. Do blue light filters confer protection against age-related macular degeneration? Prog Retin Eye Res. 2004;23:523-531.
2. Norren DV, Vos JJ. Spectral transmission of the human ocular media. Vision Res. 1974;14:1237-1244.
3. Beatty S, Koh H-H, Phil M, Henson D, Boulton M. The role of oxidative stress in the pathogenesis of age-related macular degeneration. Surv Ophthalmol. 2000;45:115-134.
4. Wu J, Seregard S, Algvere PV. Photochemical damage of the retina. Surv Ophthalmol. 2006;51:461-481.
5. Roberts JE. Ocular phototoxicity. J Photochem Photobiol B. 2001;64:136-143.
6. Bron AJ, Vrensen GFJM, Koretz J, Maraini G, Harding JJ. The ageing lens. Ophthalmologica. 2000;214:86-104.
7. Ham WT Jr, Mueller HA, Sliney DH., 1976. Retinal sensitivity to damage from short wavelength light [letter], Nature, 260, 153-155.
8. Noell WK. Possible mechanisms of photoreceptor damage by light in mammalian eyes. Vision Res. 1980;20:1163-1171.
9. Mainster MA. Violet and blue light blocking intraocular lenses: photoprotection versus photoreception. Br J Ophthalmol. 90. 2006. 784-792. Available at: Accessed November 30, 2009.
10. Mainster MA. Spectral transmittance of intraocular lenses and retinal damage from intense light sources. Am J Ophthalmol. 1978;85:167-170.
11. Mainster MA. The spectra, classification, and rationale of ultraviolet-protective intraocular lenses. Am J Ophthalmol. 1986;102:727-732.
12. Lindstrom RL, Doddi N. Ultraviolet light absorption in intraocular lenses. J Cataract Refract Surg. 1986;12:285-289.
13. Braunstein RE, Sparrow JR. A blue-blocking intraocular lens should be used in cataract surgery. Arch Ophthalmol. 2005;123:547-549.
14. Brockmann C, Schulz M, Laube T. Transmittance characteristics of ultraviolet and blue-light-filtering intraocular lenses. J Cataract Refract Surg. 2008;34:1161-1166.
15. Kernt M, Neubauer AS, Liegl R, et al. Cytoprotective effects of a blue light-filtering intraocular lens on human retinal pigment epithelium by reducing phototoxic effects on vascular endothelial growth factor-α, Bax, and Bcl-2 expression. J Cataract Refract Surg. 2009;35:354-362.
16. Lockley SW, Gooley JJ. Circadian photoreception: spotlight on the brain. Curr Biol. 16. 2006. R795-R797. Available at:,_Issue_18,_Pages_1771-1870,_R777-R815_(19_September_2006)%23tagged%23Volume%23first%3D16%23Issue%23first%3D18%23date%23(19_September_2006)%23&view=c&_gw=y&wchp=dGLbVzz-zSkzV&md5=d35a04133c4b4ce7e3ccaaa4f71db220&ie=/sdarticle.pdf. Accessed November 30, 2009.
17. Mainster MA, Sparrow JR. How much blue light should an IOL transmit? Br J Ophthalmol. 87. 2003. 1523-1529. Available at: Accessed November 30, 2009.
18. van de Kraats J, van Norren D. Sharp cutoff filters in intraocular lenses optimize the balance between light reception and light protection. J Cataract Refract Surg. 2007;33:879-887.
19. Cionni RJ, Tsai JH. Color perception with AcrySof natural and AcrySof single-piece intraocular lenses under photopic and mesopic conditions. J Cataract Refract Surg. 2006;32:236-242.
20. Mester U, Holz F, Kohnen T, Lohmann C, Tetez M. Intraindividual comparison of a blue-light filter on visual function: AF-1 (UY) versus AF-1 (UV) intraocular lens. J Cataract Refract Surg. 2008;34:608-615.
21. Shah SA, Miller KM. Explantation of an AcrySof Natural intraocular lens because of a color vision disturbance. Am J Ophthalmol. 2005;140:941-942.
22. Werner L, Mamalis N, Romaniv N, Haymore J, Haugen B, Hunger B, Stevens S. New photochromic foldable intraocular lens: preliminary study of feasibility and biocompatibility. J Cataract Refract Surg. 2006;32:1214-1221.
23. Muftuoglu O, Karel F, Duman R. Effect of a yellow intraocular lens on scotopic vision, glare disability, and blue color perception. J Cataract Refract Surg. 2007;33:658-666.
24. Rodríguez-Galietero A, Montés-Micó R, Muñoz G, Albarrán-Diego C. Comparison of contrast sensitivity and color discrimination after clear and yellow intraocular lens implantation. J Cataract Refract Surg. 2005;31:1736-1740.
25. Mangione CM, Lee PP, Pitts J, Gutierrez P, Berry S, Hays RD. Psychometric properties of the National Eye Institute Visual Function Questionnaire (NEI-VFQ); the NEI-VFQ Field Test Investigators. Arch Ophthalmol. 1998;116:1496-1504.
26. Lin I-C, Wang I-J, Lei M-S, Lin L.L.-K, Hu F-R. Improvements in vision-related quality of life with AcrySof IQ SN60WF aspherical intraocular lenses. J Cataract Refract Surg. 2008;34:1312-1317.
27. Smith VC, Pokorny J, Pass AS. Color-axis determination on the Farnsworth-Munsell 100-hue test. Am J Ophthalmol. 1985;100:176-182.
28. Bhattacharjee H, Bhattacharjee K, Medhi J. Visual performance: comparison of foldable intraocular lenses. J Cataract Refract Surg. 2006;32:451-455.
29. Wirtitsch MG, Schmidinger G, Prskavec M, Rubey M, Skorpik F, Heinze G, Findl O, Karnik N. Influence of blue-light-filtering intraocular lenses on color perception and contrast acuity. Ophthalmology. 2009;116:39-45.
30. Mainster MA. Intraocular lenses should block UV radiation and violet but not blue light. Arch Ophthalmol. 2005;123:550-555.
31. Carson D, Margrain TH, Patel A. New approach to evaluate retinal protection by intraocular lenses against age-related lipofuscin accumulation-mediated retinal phototoxicity. J Cataract Refract Surg. 2008;34:1785-1792.
32. Espindle D, Crawford B, Maxwell A, Rajagopalan K, Barnes R, Harris B, Hileman K. Quality-of-life improvements in cataract patients with bilateral blue light-filtering intraocular lenses: clinical trial. J Cataract Refract Surg. 2005;31:1952-1959.
© 2010 by Lippincott Williams & Wilkins, Inc.