Assessment of Lens Center Using Optical Coherence Tomography, Magnetic Resonance Imaging, and Photographs of the Anterior Segment of the Eye

Abstract

Purpose: To determine the nearest marker for evaluating the center of the crystalline lens using optical coherence tomography (OCT), magnetic resonance imaging (MRI), and photographs.

Methods: Optical coherence tomography scans of human eyes were obtained in vivo during femtosecond laser-assisted cataract surgery. From axial and sagittal images, the distance of the angle center (AC) and pupil center (PC) from the scanned capsule center (SCC) was calculated. From pre- and postoperative photographs, the distance of the PC and limbal center (LC) from the intraocular lens (IOL) center was calculated, and distance between each center on the lens equatorial plane was compared. After combination of pre- and postoperative images, we arranged the centers in order of distance from the IOL center. High-resolution MRI was performed in pig eyes ex vivo to confirm the exact location of the lens center relative to other centers.

Results: In human OCT scans and photographs (n = 76), the IOL center to AC distance was 0.22 ± 0.13 mm, the IOL center to SCC distance was 0.22 ± 0.12 mm, the IOL center to PC distance was 0.25 ± 0.17 mm, and the IOL center to LC distance was 0.30 ± 0.18 mm. The AC and SCC were significantly closer to the IOL center than the PC or LC. In MRI (n = 54 images), the lens center to AC distance was 0.90 ± 0.58 mm, and the lens center to PC distance was 1.53 ± 0.87 mm (Δ distance = 0.63 ± 0.69 mm, P = 0.000).

Conclusions: Optical coherence tomography, MRI, and photographs of the anterior segment revealed that the AC is the nearest marker to the center of the lens equator.

During cataract surgery, the ideal size and centration of continuous curvilinear capsulorhexis (CCC) are important. The CCC should be sufficiently large, but still smaller than the size of the optic zone of the intraocular lens (IOL).^1–3 The optimal capsulorhexis is a well-centered opening that perfectly overlaps with the IOL optic zone by 360°.⁴ The center of the CCC must be closer to the real center of the capsular bag to make the capsulorhexis as large as possible within the limits of the IOL optic zone. Decentration of the CCC relative to the lens center may cause decentration of the IOL or asymmetric capsular phimosis after cataract surgery because of uneven contractile forces arising from the asymmetric overlap between the anterior surface of the IOL and the posterior surface of the CCC margin.⁵ 

The Catalys Precision Laser System (Abbott Medical Optics, Inc., Santa Ana, CA, USA) uses axial (x-axis) and sagittal (y-axis) sectional scanned images to set up safety margins at the early planning stage when performing lens fragmentation of the anterior and posterior lens capsule. The scanned capsule is an imaginary line of the crystalline lens seen from the anterior and posterior capsule, which is visible through the dilated pupil, and is provided by the built-in algorithm of the laser system. The estimated coordinates of the angle center (AC) can be calculated based on the location of the scanned capsule center (SCC) in the equatorial plane of the lens. The system also suggests the predicted pupil center (PC), limbal center (LC), and SCC from the anterior segment in black-and-white monitor images, allowing the surgeon to select the center of the CCC prior to laser emission. 

Not all ophthalmologists can perform femtosecond laser-assisted cataract surgery (FACS) because of its cost or for other reasons; thus, SCC is not widely used. More commonly used centers such as the AC or PC can be determined using commercially available anterior segment optical coherence tomography (AS-OCT), or the LC can be determined using a surgical microscope. From postoperative photographs obtained after full dilation of the pupil, the exact center of the IOL, pupil, and limbus can be determined. Under the hypothesis that pre- and postoperative fully dilated PC should be exactly the same in the same patient, pre- and postoperative x- and y-coordinates of each PC should be fused into single coordinates based on the dilated PC. To confirm the results obtained using AS-OCT and photographs, magnetic resonance imaging (MRI) of enucleated fresh pig eyes was also used to assess the virtual lens center and the location of the other centers. 

This study was conducted to determine the center nearest to the center of the lens equator, in addition to the distance between centers such as SCC, AC, PC, and LC. 

Methods

Analysis of Human Axial and Sagittal Optical Coherence Tomography Scans

This retrospective study included 76 eyes from patients who underwent FACS at Seoul St. Mary's Hospital, Seoul, South Korea, between December 2013 and January 2015. The study was approved by the Institutional Review Board of Catholic University of Korea and was performed in accordance with the tenets of the Declaration of Helsinki. Patients were excluded if videos from the femtosecond laser planning step were not fully stored to obtain snapshots of each center, or if there were no clear photographs before postoperative week 1, or if severe decentration of the IOL occurred because of anterior capsular contracture in the postoperative 2 months. After sufficient pupil dilation was confirmed, a disposable vacuum interface (Liquid Optics; Abbott Medical Optics, Inc.) was positioned and fixed to the eye globe using a suction ring. Under such conditions, integrated scanning using spectral-domain OCT was performed. Optical coherence tomography sections of the anterior segment taken from the outer surface of the cornea to the posterior surface of the lens capsule that were wide enough to show the whole cornea were displayed in a single image. We analyzed these images using ImageJ software (National Institutes of Health, Bethesda, MD, USA). 

Because reference points from vertical OCT scans and horizontal photographs cannot be shown in a single two-dimensional image, we converted vertical reference points from these OCT scans to x- (axial) and y- (sagittal) coordinates and calculated horizontal distances between them to determine the nearest point from the lens center on the lens equatorial plane. Figure 1 provides a detailed description of the definition of the different centers, in addition to the method used to measure the distance between centers. Superior and temporal directions were converted to positive values, whereas inferior and nasal directions were converted to negative values in the location coordinates. Coordinates of the x-axis from the right eye were reversed to the equivalent negative value to analyze the left eye. 

Figure 1

 

Analysis of angle center (AC) and pupil center (PC) relative to scanned capsule center (SCC) from a vertical section of an optical coherence tomography scan of the anterior segment in a human eye in vivo. (A) The SCC was defined as the center of the line that links both ends of the scanned capsule curves. The angle center was defined as the center of the line that links both ends of the angle. The pupil center was defined as the center of the line that links both ends of pupil margin. (B) By drawing a line that runs perpendicular to the scanned capsule center line and approaches the AC and PC, the distance between the centers at the level of the scanned capsule line was calculated. The distance in the axial images is indicated on the x-axis, whereas the distance in the sagittal images is indicated on the y-axis.

Analysis of Human Preoperative and Postoperative Photographs

In FACS, a pretreatment plan that includes the location of the CCC center is displayed on the monitor. Preoperative SCC, PC, and LC are clearly overlaid on anterior segment images viewed through the liquid optics interface. Thus, the surgeon can choose the proper center to perform capsulorhexis (Fig. 2A). We analyzed these images to locate the PC and LC relative to the SCC as x- and y-coordinates. Although scale bars are not provided in these images, we could calculate the exact distance from the set diameter of the overlaid CCC. 

Figure 2

 

Analysis of centers from preoperative (A) and postoperative (C) photographs in a human eye in vivo. (A) In a planning step, a preoperative anterior photograph is displayed in the monitor of the Catalys system. The center of the purple circle is the scanned capsule center, the yellow circle center is the pupil center, and the blue circle center is the limbus center. (B) A magnified image of the quadrangular area from (A) shows the discordance between the different centers. (C) Postoperative photograph with fully dilated pupil in the same patient. Yellow circle (dilated pupil), green circle (optic margin of the intraocular lens), and blue circle (limbus) provide the exact location of each center. (D) The magnified image of the quadrangular area from (C) facilitates the calculation of the distance between centers.

Anterior segment photographs obtained from the fully dilated pupil of the same person were taken within the first week post surgery to avoid the effect of capsular contracture (Fig. 2C). These images enable the localization of the postoperative IOL center, in addition to the PC and LC relative to the IOL center, on the x- and y-coordinates. Although there are no scale bars, we used the diameter of the optic zone of the inserted IOL (i.e., 6 mm) as a ruler to calculate distances between centers. All images were analyzed based on the left eye, and x-coordinates of the right eye were converted to negative values. We hypothesized that before and after surgery, each of the centers in the fully dilated pupil is exactly the same in the same person. By overlaying the coordinates of pre- and postoperative anterior segment photographs, we converted each center based on the postoperative IOL center and recalculated the distance between the centers. We tried to determine the center nearest to the postoperative IOL center on the lens equatorial plane. 

Analysis of Pig Eyes Ex Vivo Using Magnetic Resonance Imaging

High-resolution 9.4 Tesla MRI (Bruker Biospec 94/20 USR; Bruker Biospin, Ettlingen, Germany) was performed on fresh pig eyes obtained from a local slaughterhouse. A 50-mL syringe was filled with perfluorocarbon before insertion of the pig eyes to avoid the effect of gravity on the microstructure of the eye. To obtain high-resolution images, we used an orbit coil (Bruker Biospin) on the syringe, and fixed the eyes to avoid movement. Magnetic resonance imaging was performed at 9.4 Tesla, and T2-weighted turbo spin echo images with a field of view of 40 × 50 mm (offset 2.8 mm, 1 mm) and a matrix size of 512 × 512 pixels were obtained. Fifty-four pig eyes were used, and vertical and horizontal images were obtained from each eye. 

Figure 3 provides a detailed description of the procedure used. To minimize potential bias in the calculation, MR images with artifacts or tilted eyes were excluded before analysis. 

Figure 3

 

Analysis of angle center (AC) and pupil center (PC) relative to lens center from a vertical section of magnetic resonance images in a pig eye ex vivo. (A) The lens center is the center of the line that links both sides of equators. The definition of angle and pupil center is the same as in Figure 1 (field of view = 40 × 50 mm; matrix size = 512 × 512 pixels). (B) By drawing a line that runs perpendicular to the lens center line and approaches the AC and PC, the distance from the lens center to the other centers at the horizontal lens equatorial plane can be calculated. Other images from the same pig eye rotated 90° were analyzed in a similar manner, and the offset distance is indicated on the y-axis.

Statistical Analysis

All statistical analyses were performed using SPSS software v.18.0 for Windows (SPSS, Inc., Chicago, IL, USA). Significant differences in the distance between centers were determined using paired t-tests. P < 0.05 was considered statistically significant. 

Results

The angle-to-angle and lens diameters of eyes are shown in the Table. Human eyes were analyzed after full dilation of the pupils. In the human OCT scans, the angle-to-angle diameter was not statistically different between axial (11.34 ± 0.56 mm) and sagittal (11.21 ± 0.67 mm) sectional images. This result means that the angle structure was nearly circular. 

Table

 

Basic Characteristics of Eyes

 

When pre- and postoperative anterior segment photographs were overlaid based on the dilated PC, the relative distances of the AC, PC, SCC, and LC to the IOL center were analyzed. The AC and SCC were close to the IOL center, whereas the PC and LC were farther away (Fig. 4). The AC to IOL center distance was 0.22 ± 0.13 mm, the SCC to IOL center distance was 0.22 ± 0.12 mm, the PC to IOL center distance was 0.25 ± 0.17 mm, and the LC to IOL center distance was 0.30 ± 0.18 mm. The mean IOL center to AC and IOL center to SCC distance was not significantly different (P = 0.925). The IOL center to AC and IOL center to PC distance was 0.03 ± 0.14 mm (P = 0.019), the IOL center to AC and IOL center to LC distance was 0.08 ± 0.15 mm (P = 0.000), the IOL center to SCC and IOL center to PC distance was 0.03 ± 0.12 mm (P = 0.006), and the IOL center to SCC and IOL center to LC distance was 0.08 ± 0.13 mm (P = 0.000). The IOL center to PC and IOL center to LC distance was not statistically different (P = 0.061) (Fig. 5). 

Figure 4

 

Distribution charts of each center based on the postoperative intraocular lens center. (A) Angle center. (B) Scanned capsule center. (C) Pupil center. (D) Limbus center.

Figure 5

 

Distance between centers and the postoperative intraocular lens (IOL) center. *Statistically significant.

From the MR images of the enucleated pig eyes, the diameter of the crystalline lens on the equatorial plane was determined to be 10.19 ± 0.13 mm, and there were no statistically significant differences between the long and short diameters in a given eye (P = 0.154). The angle diameter was 14.04 ± 1.03 mm, and there were no statistical differences between the long and short angle diameters (P = 0.844). (Table). 

From the lens center, the distance to the AC was 0.90 ± 0.58 mm and the distance to the PC was 1.53 ± 0.87 mm. The lens center and AC were significantly closer to each other than the lens center and the PC, which were 0.63 ± 0.69 mm apart (P = 0.000) (Fig. 6). 

Figure 6

 

Distance of angle and pupil centers from the lens center. *Statistically significant.

Discussion

Recently, the use of multifocal and toric IOLs has increased the need for obtaining well-centered round CCCs. Because an anterior capsulotomy is usually not perfectly circular, several methods to facilitate the completion of CCC have been devised.^6–8 Several studies have reported that femtosecond laser capsulotomies show better IOL centration with better circularity after surgery than manual CCCs.^9–12 However, it is difficult to create a perfectly sized anterior capsulorhexis that is also perfectly centered in relation to the lens. Some useful methods are available to locate a well-centered capsulotomy in the anterior lens capsule, such as the use of an eyecage device¹³ or algorithms that provide a SCC or PC during FACS. Clinically, surgeons locate the CCC to the center of the limbus or dilated pupil, but the eyecage device shows only an indirect center relative to the limbus. In contrast, FACS is not used worldwide for economic reasons or because of its limitations when performing surgery in small pupils or shallow anterior chambers. 

In this study, we used OCT, MRI, and anterior segment photographs to investigate which anatomical structure most closely matches the preoperative lens center and provides perfect concordance with the postoperative IOL center. Recently, several studies have attempted to perform ideal well-centered CCCs, but the results were analyzed only during the postoperative period. To our knowledge, this is the first study to determine the relative anatomical location before cataract surgery using sectional anterior segment images. 

Our study showed that the AC and SCC tended to be the centers closest to the IOL center on the lens equatorial plane. The AC and SCC were also concentrated around the IOL center without outlier points. The next nearest center was the PC, whereas the LC was widely dispersed and too far from the IOL center. In conclusion, the best landmark for crystalline lens centering is either the AC or SCC. Of these two exact landmarks, the AC is more useful because nearly all of the commercially available anterior segment OCTs or ultrasounds can provide a clear angle image. 

A limitation of our study is that the SCC and IOL center do not perfectly coincide with the crystalline lens center; they are only estimated values that can be used as substitutes for the lens center. To increase accuracy, we excluded tilted photographs, and we analyzed photographs of patients who showed maintenance of even capsular contracture around the whole optic edge after 2 to 6 months. The mean postoperative best-corrected visual acuity was logMAR 0.04 ± 0.09 (n = 76) at 2 months after cataract surgeries. Although there were no significant differences in postoperative vision according to the CCC center chosen, postoperative IOL position could be affected by decentering or tilting of the IOL or contraction of CCC. Further long-term follow-up studies for visual prognosis involving subjective vision, refractive error, and high-order aberration should be performed, because a relationship may exist between the CCC center and visual prognosis after cataract surgery. 

To improve the shortcomings of OCT scans, which cannot show retro-iris structures, we rechecked the relationships among centers in pig eyes ex vivo using MRI. In contrast to other well-established ophthalmologic imaging methods, MRI provides true anatomical proportions independent of the absorption and optical characteristics of the tissues. In a prior study by Langner et al.,¹⁴ 7.1 T MRI to assess the anterior segment of the eye was effective for discriminating microstructures such as the angle, sulcus, and lens equator. In this study, we used high-resolution 9.4 Tesla MRI with an orbit coil and set the sectional thickness to be as thin as possible to avoid artifacts and to improve image quality. The 9.4 T MRI is more powerful than 7.1 T MRI, and its use in humans is not permitted for safety reasons. In the MR images, we found the exact location of the lens equator directly, and the center closest to the lens center was the AC; the PC was significantly farther away than the AC. We tried to perform MRI in human eyes in vivo in the early stage, but finally we did not succeed because of poor image quality. Several factors such as high cost, microsaccades (i.e., involuntary eye movements), and motion artifacts from continuous breathing were significant problems in a previous human MR imaging study,¹⁵ and we experienced the same problems. However, when we performed MRI in pig eyes ex vivo rather than in human eyes, we showed that the lens center was closer to the AC than PC. 

The next question we addressed was why the AC is superior to the limbus or PC. Our results can be interpreted in two ways. First, because the limbus is not perfectly circular but rather is an ovoid-shaped structure, the horizontal corneal diameter is approximately 0.8 mm greater than the vertical diameter.¹⁶ The presence of pannus can also interfere with the estimation of the true center prior to performing capsulorhexis. Second, there is a significant shift in the location of the PC with increasing dilation, with the lateral position coming closer to the center of the cornea at its maximal dilation. A previous study showed that the PC shifts temporally as the pupil dilates, with a mean motion distance of 0.133 mm between the mesopic and photopic conditions.¹⁷ With pharmacologic dilation, the PC shifts in every direction, primarily inferotemporally, relative to the corneal reflex.¹⁸ 

The equator of the crystalline lens hangs from the ciliary body via zonular fibers, and the borderline between the anterior aspect of the iris and the corneal endothelial surface is the angle. Thus, anatomically, the angle could be a better representative of the lens equator than the pupil margin or limbus because of its geometric location. In addition, the angle has an advantage over the pupil because it is not influenced by pharmacologic dilation. A previous study reported that the corneal diameter is significantly correlated with the lens diameter (correlation coefficient = 0.711; P < 0.001),¹⁹ and there are possible close relationships between size and location for the lens, angle, and corneal structures. 

The Catalys system produces three-dimensional spectral-domain OCT images with a wavelength of 830 nm and axial and lateral resolutions of 30 and 15 μm, respectively. This technique can show every structure covered by a liquid optic interface whose clear aperture is 13.5 mm. Because the depth of field is also large, OCT scans can include the corneal epithelium to the posterior capsule of the lens. To show the center of the CCC during FACS, automated surface mapping algorithms may calculate the scanned capsule line along the anterior and posterior capsules. In contrast, anterior segment OCTs commonly used in the field of ophthalmology have transverse scan ranges of 10 to 16 mm, scan depths of 2 to 6 mm, and resolutions of 5 to 60 μm. Despite their good resolution, they are limited by their relatively smaller field of view in a single scan compared to the OCT Catalys system. Commonly used systems cannot show the scanned capsule curvature because the images that they produce lack the posterior lens capsule. Ultrasound biomicroscopes have several applications because of their probe characteristics, but in general, only the newest vesions can show the whole cornea and lens. However, despite the limitations mentioned above, all ultrasound biomicroscopes and commercially available OCTs can show the “angle structure” clearly; therefore, information regarding the AC could be applied during CCC procedures in every clinical condition. 

Placement of enucleated pig eyes in perfluorocarbon minimized the effect of gravity. Under this condition, our ex vivo study showed the true anatomical structures without deviations. In contrast, human eyes imaged in vivo under the docking state using a patient interface may be distorted by external pressure. However, the results of our study coincided with two opposite extreme conditions. Thus, our method can be used in real conditions under gravity or other pressures arising from the patient interface or nearby soft tissues. 

In conclusion, if surgeons cannot find the location of the lens equator under the surgical microscope because of the iris structure, they can determine the exact location of the lens center from the AC in OCT scans taken before surgery. This method is as useful as SCC. If surgeons perform CCC based on the AC, the postoperative capsulotomy center will be close to the real center of the lens. Other landmarks such as the pupil or limbus center are less accurate than the AC. If a surgical microscope fused with OCT is developed, surgeons will be able to determine the center of the CCC in real time by overlaying the point of the anterior capsule on the surgeon's view. 

Acknowledgments

The authors thank Hyeonjin Kim, PhD, Radiology, Biomedical sciences, Seoul National University/Hospital, for his participation in ex-vivo MR imaging in this study. 

Disclosure: Y.E. Lee, None; C.-K Joo, None