New IOL developments are focused on improving both far and intermediate vision while maintaining a functional level of near vision, as well as minimizing photopic phenomena associated with high add multifocal IOLs.1 One example of these new developments is the low add bifocal IOLs (AcrySof SV25T0; Alcon Laboratories, Inc., Fort Worth, TX, and TECNIS ZKB00; Abbott Medical Optics, Inc., Santa Ana, CA) that reduce the distance between the far and near focus generated by the IOL and therefore minimize the size of the associated halo.1,2 Another new technology is the concept of the extended depth-of-focus (EDOF) IOL (TECNIS ZXR00; Abbott Medical Optics, Inc.) that combines a diffractive echelette and an achromatic and aspheric monofocal curve on the optical zone of the IOL to extend the range of optimum focus.3,4 To date, no study has compared these two modalities of IOLs (low add vs EDOF IOLs).
The purpose of the current study was to evaluate the in vitro optical performance of EDOF and low add bifocal IOLs, including the generation of halos, and to compare it with that of high add bifocal and monofocal IOLs.
Patients and Methods
Four TECNIS IOL models (Abbott Medical Optics, Inc.) were compared in this in vitro study: monofocal (ZCB00), EDOF (ZXR00), low add bifocal (ZKB00), and high add bifocal (ZMB00) IOL. These IOLs are based on the same lens platform, with the same material, haptic configuration, and manufacturing process (Table A, available in the online version of this article). The evaluated EDOF IOL has an achromatic diffractive pattern that elongates the sharp vision of the eye and compensates for the chromatic aberration of the cornea. Specifically, it has a biconvex wavefront-designed anterior aspheric surface and a posterior achromatic diffractive surface.3,4 The low and high add bifocal IOLs have addition powers of +2.75 and +4.00 diopters (D), respectively. Both are diffractive bifocal IOLs with an anterior aspheric surface and a posterior diffractive surface. All evaluated IOLs had an optical power of 20.00 D.
Specifications for IOLs
Optical Bench System Measurement
The optical bench metrology system developed at The Catholic University of Korea was used in this study (Figure A, available in the online version of this article). This system is composed of the 1951 United States Air Force (USAF) resolution test chart, an artificial pupil, a pupil camera, a Badal optometer, a model eye, and a charge coupled device (CCD) camera (Guppy pro F503C; Allied Vision Technologies GmbH, Stadtroda, Germany). The model eye consisted of an artificial cornea (LB1014-A; Thorlabs, Newton, NJ) and a wet cell. This lens has a spherical aberration level similar to that in the healthy cornea (0.189 µm for 6-mm pupil). This modification allowed us to simulate the spherical aberration changes that occur following the implantation of aspheric IOLs with negative spherical aberration (−0.27 µm of spherical aberration for 6-mm pupil for the four IOLs evaluated) in real eyes. The front and back surfaces of the wet cell were clear flat windows. The IOL was positioned inside the wet cell filled with balanced salt solution (BSS Plus; Alcon Laboratories, Inc.) that has a refractive index similar to that of the aqueous humor (1.336). The distance between the posterior surface of the artificial cornea and the center of the IOL was set to 4.5 mm to simulate the typical anterior chamber depth of the pseudophakic eye from corneal epithelium to the center of IOL, as reported in a previous clinical study.5 Both the eye model and CCD camera were mounted on a XYZ translation stage for a precise alignment. The 1951 USAF resolution chart was used as the resolution target and positioned on the retinal image plane. The chart was illuminated by white light (Thorlabs) and the images were recorded with the CCD camera. The white light source has nearly uniform spectral energy over the visible wavelength range. This camera detected images in the defocus range from +0.50 to −4.00 D.
Schematic diagram of optical bench system. CCD = charge coupled device; IOL = intraocular lens
The effect of pupil size and IOL decentration on the image quality was measured by regulating the size of the artificial pupil and moving the wet cell 1 mm horizontally in 0.25-mm increments using the translation stage. Four different sizes of artificial pupil were used in the experiment: 2, 3, 4, and 5 mm. The pupil camera in the bench testing system was used to image the artificial pupil and IOL simultaneously. This allowed us to precisely align the IOL to the center of the artificial pupil (ie, the optical axis) and to induce different amounts of IOL decentration for the current study.
The point spread function (PSF) images were measured by replacing the USAF target with a pinhole on the same plane. The pinhole size (20 µm in diameter) served as the point source for our PSF measurement. The pinhole was used to measure the PSF images in which a small point source was required. It made us evaluate PSF in our optical bench testing. The simulation of the PSF was achieved using a point light source with a neutral density filter (Edmund Optics Inc., Barrington, NJ). It was set to the condition that the external light was focused on the foveal cone photoreceptor in the retina when the whole eye is assumed to have a focal length of 16.7 mm. PSF images were measured with a 5-mm artificial pupil and a luminance that is commonly provided by common light sources, such as sperm candle flame or T-12 fluorescent lamp (Cool white 800 mA).6 This setting was used to simulate gazing at a bright light source under mesopic conditions, such as night driving.7
The optical bench testing system was set to have the USAF target to be positioned to optical infinity. The far distance was defined as the camera position at which the image quality (ie, contrast of the bar in USAF target, corresponding to 20/20 Snellen visual acuity) was peak. The images of the 1951 USAF target were analyzed for image quality and converted to through-focus modulation transfer function (MTF). This calculation was performed assuming a spatial frequency of 14.81 cycles/degree, which is approximately equivalent to 20/40 Snellen visual acuity. MTF values for each experimental condition were displayed in the defocus curve and as a diagram with the pupil size and the level of decentration of IOLs from the center of the IOL optic zone.
PSF images were normalized for measuring dimensions of the light from the white light source. Additionally, the background noise was minimized by subtracting the background image from raw images with the IOL. The images were quantified as the diameter or intensity per one pixel. The core size was defined as the full width at half maximum of the PSF images. The peak core intensity was defined as the highest contrast at the core of the PSF image. Figure B (available in the online version of this article) illustrates the method for analysis in terms of halos. The relative halo intensity was defined as the percentage of the energy distributed to the halo compared to total energy (outside of the red circle in Figure BB). The core size for each IOL was fixed to that estimated from the PSF image at far distance.
(A) The point spread function (PSF) image was converted to a logarithmic scale for the analysis of halo. (B) The relative halo intensity was calculated as the percentage of the energy distributed to the halo (outside of red circle) compared to total energy. The intensity profile in a logarithmic scale (C) for the converted image (A) of the PSF image was obtained. The core size of the PSF images was calculated with the definition of full width at half maximum (FWHM).
Descriptive statistical analysis was performed using SPSS software (version 18.0; SPSS ,Inc., Chicago, IL). The Spearman correlation test was used for the analysis to confirm the correlation between the MTF for 1951 USAF target images and the halo-related parameters for PSF images according to the defocus diopter. A P value less than .05 was considered statistically significant.
Image Quality Analysis
The image quality for far distance was best in the range from +0.50 to −4.00 D of defocus for all four different types of IOL. The monofocal IOL had the best far image, followed by the EDOF, low add bifocal, and high add bifocal IOLs (Figure 1A). The EDOF IOL showed stable image quality at intermediate distances (1.00 D [100 cm], 1.50 D [66 cm], and 2.00 D [50 cm] of defocus) (Figure 1B). The near images corresponding to −2.10 D (46.7 cm) and −3.00 D (33.0 cm) of defocus were better with low and high add bifocal IOLs, respectively. The low add bifocal IOL showed better image quality than the high add bifocal IOL, but a lower image quality compared to the EDOF IOL at a defocus of −1.00 D (100 cm) and −1.50 D (66 cm).
Representative 1951 United States Air Force (USAF) target images for the four different intraocular lenses measured at (A) far distance and (B) between intermediate and near distance with 0.50 diopters (D) of defocus. All images were measured using a 3-mm artificial pupil. The far distance was defined as the camera position at which the image quality of the bar in the USAF target, corresponding to 20/20 Snellen visual acuity, was peak. EDOF = extended depth of focus
Through-Focus MTF Analysis
The image quality at far distance of the EDOF IOL was the most comparable to that of the monofocal IOL among the three presbyopia-correcting IOLs and showed a gradual decrease in MTF at a defocus range between 0.00 and −1.70 D (Figure 2). Low and high add bifocal IOLs provided two peaks of maximum optical quality corresponding to the far and near focus (2.10 and 3.30 D of defocus, respectively) in the MTF curve.
Through-focus modulation transfer function (MTF) of the four different intraocular lenses (IOLs) evaluated. This measurement was done for a 3-mm artificial pupil and a spatial frequency of 14.81 cycles/degree. EDOF = extended depth of focus
The MTF decreased by 3.6% when changing from a pupil size of 2 to 5 mm with the monofocal IOL. In contrast, the decrease in MTF associated with this pupil size change was 4.9%, 5.7%, and 6.2% with the EDOF, low add bifocal, and high add bifocal IOLs, respectively (Figure 3A). IOL decentration of 1 mm generated an MTF decrease of 35.1%, 27.8%, 39.1%, and 34.8% with the monofocal, EDOF, low add bifocal, and high add bifocal IOLs, respectively (Figure 3B).
Analysis of the modulation transfer function (MTF) for the (A) pupil size and (B) different amounts of decentration of the intraocular lens with respect to the center of the intraocular lens optic. Measurements were performed for a 3-mm artificial pupil and spatial frequency of 14.81 cycles/ degree. EDOF = extended depth of focus
The through-focus images, including PSF images, were converted to a logarithmic scale (Figure 4) for a more detailed perception of glare or halo pattern. Glare is typically produced by randomly distributed light scatter causing an increase in background noise in the retinal images. Typical sources of glare include cataract or IOL edge.8,9 In contrast, halo is defined as more regular patterns around an object mainly caused by optical designs of IOLs such as diffractive and refractive rings included in multifocal IOLs. As shown in PSF images at 0.00 D of defocus in Figure 4, the glare pattern was dominant in the PSF images of monofocal and EDOF IOLs. In contrast, the halo was dominant in the PSF images of low add and high add bifocal IOLs.
Point spread function (PSF) images for four different intraocular lenses were captured with a 20-µm diameter pinhole illuminated by the white light source. The original PSF images were converted to a logarithmic scale for a clear distinction of the halo. The mesopic condition was simulated with 5-mm artificial pupil in the optical bench test. EDOF = extended depth of focus; D = diopters
The pattern of the through-focus peak core intensity in the PSF images of the four IOLs (Figure 5A) had a statistically significant positive correlation with that of the through-focus MTF described in Figure 2 (Table 1). The monofocal IOL showed the highest peak core intensity for far distance, the EDOF IOL the highest value for 1.00 D of defocus (intermediate distance), the low add bifocal IOL for 2.00 D of defocus (intermediate distance), and the high add bifocal IOL for near distance. The relative halo intensity of the EDOF and monofocal IOLs showed a consistent increase with the level of defocus (Figure 5B). Those of the low and high add bifocal IOLs were stable at intermediate (1.00 and 2.00 D of defocus) and near (2.50 and 3.00 D of defocus) distances, respectively. The relative halo intensity and the MTF were negatively correlated within a defocus diopter range from 0.00 to −3.00 D for the four IOLs evaluated (Table 1).
The halo that was found in the point spread function (PSF) images was evaluated by the (A) peak core intensity and (B) relative halo intensity (B) for four different intraocular lenses (IOLs). The relative halo intensity was calculated as the percentage of the energy distributed to the halo compared to total energy. The core size for each IOL was fixed to that estimated from the PSF image at far distance. EDOF = extended depth of focus; D = diopters
Correlation Between the Modulation Transfer Function for 1951 USAF Target Images and the Halo-Related Parameters for Point Spread Function Images According to the Defocus Diopter
In the current optical bench study, we evaluated the through-focus optical performance in terms of MTF and halo in the PSF image for an EDOF IOL compared to low add bifocal, high add bifocal, and monofocal IOLs. We used a specifically developed optical bench metrology system. An aspheric doublet was recommended as the artificial cornea based on International Organization for Standardization (ISO) 11979-2. However, it has been known that the typical human cornea has positive spherical aberration, so we instead used a spherical biconvex lens that induces spherical aberration similar to what the typical human cornea has for a 6-mm pupil diameter. This provided a more accurate evaluation of recent aspheric IOLs designed to compensate for some of the corneal positive spherical aberration.10,11
Halo could be generated from especially diffractive presbyopia-correcting IOLs, which make more than two images at different focal lengths and produce the degradation of visual performance.12 According to the PSF images, the relative halo intensity showed a negative correlation with MTF in the through-focus analysis for all IOLs in the current study. The halo may be a major contributor to the image quality degradation that has been found in the MTF analysis. Van der Mooren et al.13 reported the effect of stray retinal light on the degradation of visual performance by showing a positive correlation with halo size, luminance threshold, and contrast sensitivity.
We hypothesize that the blur around the focused image is caused by the halo formation related to the optic design of multifocal IOLs. In the optic design of multifocal diffractive IOLs, the width of one ring is related to the addition power for near or intermediate focus. Indeed, diffractive IOLs have different numbers of rings according to the add power provided.14 Some studies have demonstrated that positive dysphotopsia with diffractive presbyopia-correcting IOLs could be attributed to the ring pattern optic design for the multifocality along the optical axis, the airy pattern from pupil diffraction, and the saturation from the light source.4,15,16 Although Luttrull et al.17 described a concentric ring-related dysphotopsia in patients implanted with diffractive multifocal IOLs, there has been minimal research about positive dysphotopsia with multifocal IOLs. Gatinel and Loicq18 stated that the halos generated by multifocal IOLs are the result of the combination of dispersing rays and pupil diffraction. They also suggested that this halo may be related to a decreased image quality.18 In our study, we measured the PSF images using a pinhole to check the halo around the focused image. Gatinel and Loicq18 and Carson et al.15 also used a pinhole for their simulations and Vega et al.1 used a slit pattern to simulate the dysphotopsia caused by each IOL in their study.
EDOF and low add bifocal IOLs could be categorized as an EDOF IOL if the add power is sufficiently small, which produces a continuous extension of depth of focus instead of the bimodal defocus curve found from typical bifocal IOLs. In other words, defocus curve from far to intermediate distance with the low add bifocal IOL can show relatively small changes as seen in EDOF IOLs. The EDOF IOL has fewer diffractive rings than low add bifocal IOLs. For those reasons, the halo pattern in the PSF images of the EDOF IOL was similar to that found in the monofocal IOL. This is consistent with the clinical data reported in the Concerto Study Group, revealing that the same EDOF IOL as evaluated in our optical bench experience has a clinical advantage in terms of photopic phenomena.19
In the through-focus MTF curve analysis, we found that bifocal IOLs showed a bimodal pattern, with two clear peaks for far and near distance conditions, and a valley corresponding to intermediate distance. Although trifocal IOLs were developed to avoid this valley in intermediate vision, previous studies have also shown the presence of valleys between the three peaks of maximum optical quality corresponding to far, intermediate, and near distance.7,20–24 However, from a theoretical point of view, the EDOF IOL is more restrictive in terms of considering visual acuity for all distances because it causes a plateau between far and intermediate distance in the through-focus MTF curve. Specifically, an increased depth of focus ensures that the intermediate visual quality is higher than a certain level, which can be somewhat lower than that at far distance.25 Hence, the EDOF IOL can be defined as one peak of maximum optical quality for distance vision that gradually decreases, without generating a valley between far and intermediate distances. This progressive decrease of optical quality from far to intermediate distances is what was observed with the EDOF IOL evaluated in our optical bench experiment. Furthermore, better intermediate distance optical performance was observed with the low add bifocal IOL compared to the high add bifocal IOL, which is also consistent with the clinical data, showing better visual performance at intermediate vision with low add bifocal IOLs compared to high add bifocal IOLs.26,27 PanOptix (Alcon Laboratories, Inc.) and Lentis Mplus (Oculentis GmbH, Berlin, Germany) IOLs use different mechanisms to extend the depth of focus. The PanOptix IOL was designed to have add powers at far, intermediate, and near distances,28 whereas the Lentis Mplus IOL produces asymmetric transition power between the far and near zone on the IOL optic, increasing higher order aberrations such as coma.29 These IOLs can improve intermediate image quality, but it is important to note that distance quality can be compromised more compared to a bifocal IOL.
The depth of focus can also be extended due to clinical reasons, such as the level of aberration of the cornea or IOL or in relation to the pupil size in pseudophakia,17,18 showing a clear example of extended depth of focus generated by the induction of spherical aberrations. If spherical aberration is added to a mono-focal IOL, the image quality at far distance would be decreased but the depth of focus would be extended. Negatively and positively increased spherical aberration extends depth of focus toward the near focus and focus beyond the infinity, respectively. Zheleznyak et al.30,31 confirmed that spherical aberration could improve through-focus visual acuity and depth of focus, which has been found in hyperopic refractive surgery32 or modified monovision in cataract surgery.31,33 Domínguez-Vicent et al.34 reported in an optical bench experiment that with the Mini WELL Ready progressive multifocal IOL (SIFI, Catania, Italy), which is a lens inducing controlled levels of spherical aberration, a significant defocus tolerance at distance and near vision could be achieved.
We evaluated an EDOF IOL that positively increases depth of focus using exclusively diffractive optics.18 The slightly decreased visual quality at the peak for far distance with this diffractive EDOF IOL was outweighed by the compensation of corneal chromatic and spherical aberration.3 This EDOF IOL has shown a better and more stable image quality between far and intermediate distance (−1.50 D of defocus) compared to low add and high add bifocal IOLs. Our results are consistent with those found in earlier optical bench studies. Specifically, Esteve-Taboada et al.4 reported that the EDOF IOL was less vergence dependent than existing trifocal IOLs under dim conditions. Likewise, the EDOF IOL has shown disadvantages at near distance as reported in previous studies.4,18,19 Additionally, other IOLs with asphericity around zero could increase depth of focus more effectively than the EDOF IOL because corneal spherical aberration remains uncorrected. Which IOL would be better for a patient strongly depends on the patient's needs and expectations.
Finally, the effects of pupil size and IOL centration were also investigated. Our results are comparable to those obtained in previous studies, with a degradation of MTF with increasing pupil size.15,35–37 Furthermore, when the IOL was not centered on the visual axis, a decrease in MTF was observed for all IOLs, including the monofocal IOL, and this result was comparable to that reported in previous studies.38,39
Although optical bench testing provides useful insights to enable clinicians to better understand patients’ visual outcomes, it can only predict the visual quality monocularly. However, the characteristics of individual IOLs found in this study can help surgeons to apply a binocular approach (eg, mix and match) to further improve through-focus visual quality binocularly. In addition, another limitation was that the results of this study included the effect of saturation of the light source. The size of the light source used in our study was not perfectly punctual because it should theoretically be for the measurement of PSF. Finally, the intensity depth of the CCD camera was 8 bit, which, converted into numerical intensity values, is related to a range from 0 to 255. A camera with a higher range of contrast detection than 8 bit might have delivered more accurate results.
Presbyopia-correcting IOLs show a specific optical behavior in terms of through-focus MTF, PSF, and halo formation according to their optic design. The EDOF IOL seems to be a preferable option for patients with cataract requiring high levels of visual acuity and low levels of positive dysphotopsia due to halo at both far and intermediate distance after surgery. However, it should be considered that near vision is more limited than with high add bifocal IOLs and the halo at intermediate distance may cause some level of unexpected disturbance under mesopic conditions, comparable to that of diffractive bifocal multifocal IOLs.