Dynamic Purkinje Meter as a Tool for Intraocular Lens Position Measurement

Abstract

Due to the increasing demands of today’s society on visual quality and patient comfort, and due to the growing interest in the implantation of new and more complex intraocular lens (IOL) designs, determining the IOL position occupies an important position in current ophthalmological practice. The dynamic Purkinje meter combines the construction of static Purkinje meters, presented in recent years, with dynamic examination of the IOL position according to the optical axis of the IOL.

1. Introduction

A cataract is a clouding of the crystalline lens of the eye. In most cases, it is an age-related, physiological process causing a progressive deterioration of visual acuity, which can lead to complete vision loss. The prevalence of cataracts increases with age—Liu Yu-Chi et al. state that cataracts affect 3.9% of the population between 55 and 64 years, whereas in the population older than 80 years, this figure increases to even 92.6% [1]. Currently, there is no effective method of prevention or pharmacological treatment of cataracts, so the only effective solution for cataracts is surgery [2]. This is the most frequently performed surgical procedure, with over 26 million cataract surgeries performed worldwide every year [2]. The goal of the procedure is to remove the clouded lens and replace it with an artificial intraocular lens (IOL), which is implanted in the remaining part of the lens capsule.

Nowadays, cataract surgery is no longer just a process of removing a clouded lens but is becoming an integral part of refractive surgery. A refractive error can be corrected by implanting a precisely calculated IOL. It is also possible to adapt the choice of the IOL design to correct selected higher-order aberrations [3]. However, to obtain a sharp retinal image, a correct (i.e., as intended during its calculation) and stable IOL position in the lens capsule is necessary. If the lens is displaced relative to the optical axis of the eye, the retinal image quality degrades, which negatively affects the resulting visual perception. The degree of influence that the real IOL position has on the visual quality depends on deviations from the planned IOL position (i.e., IOL decentration and tilt) and on the IOL type used [4]. It is proven that IOLs with a more complex design are more sensitive to changes in IOL position—this applies to lenses correcting spherical aberration of the cornea (aspheric designs) and lenses providing greater depth of field (EDOF and multifocal lenses) [3]. As the implantation of IOLs with these designs is more and more frequent, the interest in measuring IOL position continues to grow. The interest has been growing for several reasons: explanation and solving of refractive error, refinement of the IOL calculation before surgery on the second eye, study of new IOL designs and optimization of surgical procedures (capsulorhexis, placement of IOL haptics).

Methods for measuring the IOL position used nowadays in clinical practice can be based on ultrasound biomicroscopy (UBM), optical coherence tomography (OCT), the Scheimpflug principle and the analysis of the position of Purkinje images [5,6,7,8,9]. Apart from so-called Purkinje meters, these methods are associated with expensive instrumentation and, in most cases, complex processing of the results.

Systems based on the analysis of Purkinje images [10] determine the IOL position by the localization of light reflexes reflected from four optical surfaces in the eye. There are four light reflexes: from the anterior and posterior surface of the cornea (first and second Purkinje images, P1 and P2, respectively) and from the anterior and posterior surface of the lens (third and fourth Purkinje images, P3 and P4, respectively). P2 is clinically indistinguishable as it is overlapped by P1—due to the small thickness of the cornea and also due to the small difference between the refractive index of the cornea and the aqueous humor.

In 1988, Phillips et al. [11] conducted a study with 14 pseudophakic patients with an implanted IOL, in which a patient focused his or her gaze on an object coaxial with the camera axis or located at a predefined angle to the camera axis. They showed that P3 and P4 displacements were dependent on fixation angle, IOL tilt and IOL decentration and that P1 displacement was a function of fixation angle only.

In 1990, Guyton et al. [12] described a quick method (rather of an indicative nature) for detecting IOL decentration and IOL tilt without the need for pupil dilation. The method is based on determining the optical axis of the implanted lens. By moving the light source in front of the patient’s eye, the examiner evokes Purkinje images and tries to achieve a superposition of P3 and P4. The tilt of the lens corresponds to the angular difference between the beam of the source and the axis of vision at the moment of P3+P4 superposition. Decentration can be determined as the distance between the optical axis of the lens and the center of the pupil.

Based on Phillips’ and Guyton’s findings, in 1993, Cendelin et al. [13] described a dynamic Purkinje meter they had developed—this device had a dynamic arrangement with a movable fixation target, and it allowed the assessment of a large range of IOL decentration and tilt. The Purkinje meters mentioned in the literature in the last 20 years [6,14,15,16] have a static arrangement with a fixed position of one or more fixation stimuli, which limits the measurement of the position of the IOL with a greater degree of decentration or tilting.

The purpose of this paper is to present an upgraded dynamic Purkinje meter whose design is based on the device developed and described in 1993 by Cendelin et al. [13], mentioned above.

2. Materials and Methods

2.1. Original Concept

Cendelin, Korynta and Bok developed a method for IOL position measurement in vivo, and they presented it in 1993 at the 6th Congress of the German-speaking Society for Intraocular Lens Implantation [13]. The development of the device was based on the work of Phillips et al.: Measurement of intraocular lens decentration and tilt in vivo [11], mentioned above. In this study, the authors showed that P3 and P4 displacements were dependent on fixation angle, IOL tilt and IOL decentration and thus showed that it was possible to determine the IOL position in the eye from the position of P3 and P4. The method of so-called Purkinje-metry used direct measurement of Purkinje image positions (without photography) and computer analysis for determining the IOL position.

Cendelin et al. [13] assembled an instrument for determining the position of the light source as an attachment for the Opton slit lamp (Figure 1). The attachment was attached to the front of the slit lamp so that it was clearly visible to the patient. The patient fixed the central “target” LED (light emitting diode, D1). Four coaxially arranged LEDs (D2), placed around the diode D1, formed a coaxial P1. All the diodes D1 and D2 had a fixed position during the entire measurement. The LED light source with an adjustable position (LQ, Figure 1 and Figure 2) for evoking P3 and P4 in the eye was placed on a rotating arm; i.e., the source was movable in a plane perpendicular to the optical axis of the device. It was necessary to make P3 relatively centered with P1 (i.e., to place P3 in the center of the reflex of the four LEDs, D2). Subsequently, P4 had to be relatively centered with P1 (i.e., P4 had to be set in the center of the reflex of the four LEDs, D2). The relative centration of the Purkinje images was performed by adjusting the position of the light source (LQ) on the rotating arm. The position of the adjustable light source (LQ) relative to the optical axis of the device could be clearly determined—it was given by the decentration size relative to the device axis and by the decentration meridian. The examiner observed the situation through the binocular oculars of the slit lamp microscope.

For such an arrangement, it was necessary to create a program for processing the measured values. Therefore, the concept included the creation of a computer program for calculating the IOL position from the values of the Purkinje image positions and the optical parameters of the eye. This program was a general program for calculating an optical system using the ray-tracing method.

2.2. Upgraded Device

A new dynamic Purkinje meter follows the main principles and builds on the concept of the original dynamic Purkinje meter described in the previous section.

The dynamic Purkinje meter produces Purkinje images in the eye. Their relative position is measured and subsequently computer-processed or converted to the IOL position data. The mechanical device is mounted on a slit lamp base. To achieve a sharp image of the observed field, the body of the unit is movable to focus accurately. The working distance of 40 cm is ensured by locking the focus of the camera objective. However, the camera objective must be movable to adapt to different shapes of the face and to the position of the eyes of the examined persons. The Purkinje meter includes headrests for the forehead and chin of the patient to ensure minimal movement during the examination. An occlusion to cover the non-examined eye is attached to the headrest. The dominant component of the Purkinje meter is a rotatable disc of 44 cm in diameter with lighting and control elements, which are described below. The largest possible diameter that maintained good handling with the device was chosen. In the central part of the disc around the camera porthole, there is a circle formed by 12 infrared (IR) LEDs used to evoke Purkinje images in the examined eye. The circle diameter was chosen as the smallest possible in relation to the camera objective. The central part of the disc around the camera porthole has a fixed position so that the evoked reflexes are stable. The rest of the dominant disc is rotatable, as mentioned. The Purkinje images are captured by a camera which is attached to the slit lamp body instead of a microscope. The camera used is a ZWO ASI224MC Color Camera 1.2M (ZWO, Suzhou, China) with high sensitivity in the near-IR region and a Computar 55 mm TEC-55 telecentric lens. The examiner follows the camera image on the computer monitor.

From the patient’s side, the following elements can be seen on the disc (Figure 3):

• A red-light fixation target (A) placed at a 40 cm distance in front of the patient’s eye—the examiner moves with the target, and the patient has to follow it throughout the examination (the movement of the fixation target is carried out laterally using a mini lever attached to the ruler, and at the same time the entire disk is turned by hand);
• Twelve white LEDs (B) arranged in a circle at the outer periphery of the disc, for better illumination of the measured field;
• Twelve IR LEDs (C) arranged in a circle around the central window, which generate Purkinje images and have a fixed position (i.e., the inner disc with the IR LEDs does not rotate together with the disc);
• A camera lens (D) in the central window of the disc.

From the examiner’s side, the following elements can be seen on the disc (Figure 4):

• A protractor (E) with a range of 360°, which shows the degree of rotation of the disc;
• A bar with a ruler (F) with a scale from −22 to +22 cm; the bar is movable, and at its origin (i.e., 0 cm) there is a fixation target placed from the other side of the disc;
• A control panel with lever switches (G) for setting lighting elements, described below.

The Purkinje meter contains a control panel with four lever switches (marked as G in Figure 4), which allows setting the mode of white and IR LEDs (two levers for each type of diode). The white LEDs can have different modes: “off”, “on with lower light intensity” or “on with higher light intensity”. The mode is set as needed during a specific examination so that the Purkinje images evoked in the eye have the highest possible quality on the video recording. IR LEDs that elicit Purkinje images can be triggered in full-circle mode for more accurate centering of Purkinje images or in half circle mode for easier identification of Purkinje images. The arrangement of IR LEDs evoking the Purkinje images was inspired by a static Purkinje meter designed and developed by Tabernero [14,15]. IR LEDs are available in “off”, “full circle” or “semicircle” modes or “inverse semicircles” (Figure 5).

2.3. Measurement Procedure Protocol

The developed methodology of examination on the dynamic Purkinje meter is as follows:

• The examination is carried out in a darkened room.
• The examiner instructs the patient about the course of the examination—the patient has both eyes open throughout the examination (one of which is covered), can blink as needed and follows the red-light fixation target, which the examiner moves during the measurement.
• The patient sits on the chair and rests his or her chin and forehead on the headrests.
• The examiner adjusts the height of the table and chinrest so that the mark on the headrest is at the level of the patient’s outer corner of the eye; the patient’s non-examined eye is covered with an occlusion.
• The examiner checks the setting of the fixation stimulus to the “0” position on the bar with the ruler and the protractor and turns on the red light of the fixation target with the lever switch placed directly at the fixation light.
• Using a control panel with lever switches, the examiner turns on the IR LEDs in full-circle mode.
• The examiner runs a live eye-scanning program on the computer. Throughout the examination, he or she watches the evoked Purkinje images on the computer monitor.
• If the image of the scanned eye on the monitor is too dark, the examined field can be illuminated by turning on the white LEDs around the disk using a control panel with lever switches.
• The patient is instructed to look straight ahead at the fixation target in the zero position, and the examiner records a video of the Purkinje images in whole-circle mode.
• The examiner switches the IR LEDs to the semicircle mode, identifies the Purkinje images and centers them—P3 is the larger of the upper semicircles and P4 is inverted, i.e., the lower semicircle.
• The patient is then asked to observe the fixation target, which the examiner moves until P3 and P4 are centered. The examiner performs the movement of the fixation stimulus by combining (a) the sliding movement of the fixation stimulus along the bar with the ruler and (b) turning the entire disc.
• To verify the centration of P3 and P4, the examiner switches the mode back to whole circles.
• At the moment of the superposition of P3 and P4, the examiner turns off the program for live eye scanning. The examiner saves the video recording in the computer memory under the patient’s anonymized identification number, labeled right/left eye.
• The examiner records the measured values: –The value of the lateral displacement (cm) from the bar with a ruler; –The value of the angular displacement of the fixation stimulus (degrees) on the protractor.
• The examiner turns off the fixation light and IR LEDs using the lever switches on the fixation stimulus and on the control panel.

2.4. Calculation of IOL Tilt

The IOL tilt is characterized by the tilt magnitude (degrees) and the tilt orientation (degrees). The tilt direction corresponds to the angle which is measured using the protractor on the Purkinje meter at the moment of the P3 and P4 superposition (Figure 6). To calculate the tilt magnitude, it is necessary to know the examination distance (40 cm constantly) and the fixation target position at the moment of P3 and P4 superposition. The fixation target position (FP, in centimeters) can be read from the scale bar during the measurement. Based on the knowledge of the parameters mentioned above, the tilt magnitude (α, in radians) is described by the following equation of a tangent function:

$$\mathrm{tg}\alpha =\frac{FP}{40}$$

(1)

An angular shift of the fixation target (FP_A) can be read from the protractor during the measurement. The tilt direction (β, degrees) is calculated using the following equation:

β = FP_A + 180.

(2)

2.5. Calculation of IOL Decentration

The IOL decentration is characterized by the decentration magnitude (millimeters) and the decentration orientation (degrees). The decentration is given relative to the visual axis, i.e., relative to the P1 position during the patient’s on-axis fixation.

A custom-developed software is used for the IOL decentration evaluation. Using this software, it is possible to calculate the relative positions of any interlaced circles through embedded photographs or illustrations. Video screenshots of the eye with the Purkinje reflexes captured during the measurement are inserted in the software; the screenshot captured during the on-axis fixation and the one captured during the P3 and P4 superposition are overlayed (Figure 7). For each subject, it is necessary to calibrate the software to the value of the patient’s corneal diameter (“white to white”); after that, it is possible to determine the relative position of (a) the P1 center during the on-axis fixation and (b) the center of P3+P4 superposition during off-axis fixation. Their relative position corresponds to the IOL decentration value (both magnitude and direction are obtained).

2.6. Methods

The method was verified by a study. This study included patients who underwent cataract surgery on one or both eyes and had no other significant ophthalmologic comorbidities. Patients with corneal opacities or unable to focus on the fixation target during the measurement (for example due to the presence of nystagmus) were not included in the study. All patients had a 1-piece hydrophobic acrylic IOL (AAB00 [17], Johnson & Johnson, Santa Ana, CA, USA) implanted in their eye(s). The dimensions of this IOL given by the manufacturer are 13 mm for the entire diameter including haptics and 6 mm for the diameter of the optical part only. All the operations were performed by the same surgeon.

This work was performed in accordance with the Declaration of Helsinki. The study was approved by the Ethics Committee for Multi-Centric Clinical Trials of the University Hospital Motol and Second Faculty of Medicine, Charles University in Prague, Czech Republic (reference No. EK—1019/23). All patients were informed about the nature of the study and provided written informed consent to participate in this study.

The IOL position was measured in all the patients using the developed dynamic Purkinje meter. The tilt and decentration of the IOL were measured using the measurement procedure protocol described above in Section 2.3. The IOL tilt and decentration calculations were performed following the instructions described in Section 2.4 and Section 2.5.

The values of IOL tilt and decentration measured with the dynamic Purkinje meter were evaluated using mean values and standard deviations.

3. Results

The study included 24 pseudophakic eyes (15 right and 9 left eyes) of 16 patients (8 men and 8 women). The mean age was 73 ± 10 years (range 43 to 86 years). The mean (spherical) power of the IOL was 22 D ± 2 D. IOL position measurements were performed from 5 days to 30 months after the cataract surgery; the mean was 161 days.

The mean IOL tilt magnitude was 5.7° ± 1.5°, and the mean IOL tilt direction was 265° ± 71°. The mean IOL decentration magnitude was 0.44 mm ± 0.20 mm, and the mean IOL decentration direction was 214° ± 102°. The values of IOL tilt and decentration (magnitude and direction) measured using the dynamic Purkinje meter and calculated mean values with standard deviations are listed in

Table 1. The values of IOL tilt and decentration (magnitude and direction) measured using the dynamic Purkinje meter and calculated mean values with standard deviations.

Subject	IOL Tilt		IOL Decentration
	Magnitude (°)	Direction (°)	Magnitude (mm)	Direction (°)
S1	5.0	343	0.35	353
S2	9.1	194	0.60	182
S3	4.4	204	0.36	139
S4	5.7	348	0.71	332
S5	6.1	222	0.33	120
S6	3.7	362	0.52	28
S7	5.0	190	0.19	293
S8	5.7	240	1.10	304
S9	4.4	185	0.32	156
S10	6.0	193	0.47	212
S11	6.3	341	0.70	306
S12	6.3	204	0.28	167
S13	6.7	346	0.26	346
S14	5.9	184	0.54	167
S15	1.1	319	0.32	225
S16	8.1	349	0.46	339
S17	5.3	208	0.39	177
S18	5.9	331	0.41	345
S19	6.6	196	0.40	174
S20	6.7	339	0.21	53
S21	3.9	328	0.67	21
S22	5.6	191	0.28	167
S23	6.7	199	0.39	196
S24	7.4	340	0.41	345
mean	5.7	265	0.44	214
SD	1.5	71	0.20	102

.

In the patient marked as S8, the IOL decentration magnitude value was found to be extremely high in comparison to the values of the other subjects. This patient (S8) was a 69-year-old man who underwent cataract surgery on the right eye. The power of the implanted IOL was +21 D. All the IOL position measurements of this subject were performed under mydriatic conditions and during the same control, 5 days postoperatively. The IOL position values of this “outlier” measured using the Purkinje meter were verified using two methods:

Firstly, the IOL position was determined using a commercially available anterior segment optical coherence tomography (AS OCT) instrument, CASIA2 (Tomey Corp., Nagoya, Japan); those values can be found in the Post-op Cataract protocol of the device. The trace lines were checked on both surfaces of the IOL before analysis was conducted, and if needed, the outlines of the IOL were drawn manually using the edit trace button [18]. The automatic detection of IOL boundaries was checked in all 16 cross-sections. Then, the device itself calculated the IOL decentration and tilt based on the data achieved in the 16 different cross-sections of the anterior segment of the eye.

Secondly, the real IOL center decentration relative to the P1 position during the patient’s on-axis fixation was determined. This procedure followed the one that had been described above (Section 2.5): in the screenshot with the Purkinje image positions captured during the patient’s on-axis fixation, the IOL edge was interlaced with a circle (Figure 8). The center of this circle corresponds to the real IOL center. The real IOL center position (blue in Figure 8) was compared with the P1 center position (yellow in Figure 8) from the same image, and thus, the real value of the IOL decentration magnitude was obtained. This method is applicable only in eyes with sufficiently large pupils to see the IOL edge.

The resulting IOL position values of the “outlier” achieved by the different methods and the differences between them are shown in

Table 2. IOL decentration values measured using different methods and the absolute value of the difference between them.

Subject S8	Tilt		Decentration
	Magnitude	Direction	Magnitude	Direction
Purkinje meter	5.7°	240°	1.10 mm	304°
CASIA2	5.7°	239°	0.66 mm	314°
Acc. 1 to IOL center	x	x	0.94 mm	305°
Difference Pm-CASIA2 2a	0°	1°	0.44 mm	10°
Difference Pm-IOL center 2b	x	x	0.16 mm	1°

¹ According; ² absolute value of the difference (a) between the values measured by Purkinje meter and by CASIA2, (b) between the values measured by Purkinje meter and according to IOL center.

.

4. Discussion

We developed a new dynamic Purkinje meter based on the dynamic Purkinje meter developed by Cendelin et al. [13]. The main upgrades made to the original device are as follows:

(1)

A movable fixation target instead of a static fixation target;

(2)

A static and circular (and optionally semicircular) light source evoking Purkinje images in the patient’s eye instead of a movable Purkinje image source;

(3)

A coaxial objective instead of binoculars;

(4)

A longer examination distance, adjustable by focusing the objective lens (not present in the original concept);

(5)

A ray-tracing program is not necessary in the upgraded concept.

All these upgrades allow an examiner to determine the IOL position more precisely and easily.

The (semi)circular design of the light source (originally used by Tabernero et al. [14,15]) has several advantages over a point light source. Thanks to the width of the source, it is possible to localize even Purkinje images that are partially hidden behind the iris. In addition, the option of a circular or a semicircular mode allows an examiner to identify particular Purkinje images more easily, especially P4, which is (as the only one) inverted. The circular shape of the source is necessary for the dynamic Purkinje meter due to the performed superposition of P3+P4.

The device arrangement and the design of the Purkinje image source influence the need for pupil dilation for IOL position measurement. In the systems with a static arrangement (even in those with a semicircular Purkinje image source and in eyes with a dilated pupil), identifying the exact relative position of Purkinje images is influenced by the design of the IOL. In the case of the on-axis fixation examination, IOLs with one (anterior or posterior) flatter surface produce a large P3 or P4. Such a large image is visible only partially or is not visible at all; i.e., the reflex is partially or completely covered by the iris. This complication occurs due to a static arrangement of the Purkinje meter.

The advantage of the dynamic arrangement is mostly in the possibility of setting a defined change in the fixation target position associated with a dynamic evaluation of the relative position of the Purkinje images. Thus, the method allows the exclusion of mistakes of a “random” static image and the measurement of a large range of changes in the IOL position even in eyes with relatively narrow pupils.

The method was verified by a study, in which 24 eyes of 16 patients with an implanted monofocal spherical IOL were included. The resulting mean values of the IOL position were 5.7° ± 1.5° for the IOL tilt magnitude and 0.44 mm ± 0.20 mm for the IOL decentration magnitude. All the values are given relative to the visual axis.

For comparison, some other published results (mean value ± standard deviation) are mentioned, supplemented by reference structure and measurement method: tilt 5.6° ± 1.6° and decentration 0.18 mm ± 0.12 mm (Xiao et al. [9], AS OCT, corneal topographic axis), tilt 2.9° ± 0.9° and decentration 0.56 mm ± 0.31 mm (Wang et al. [19], AS OCT Visante, pupillary plane), tilt 4.4° ± 2.5° and decentration 0.44 mm ± 0.19 mm (Maedel et al. [20], Tabernero’s Purkinje meter, pupillary axis), tilt 9.20° ± 6.96° and decentration 0.74 mm ± 0.91 mm (Maedel et al. [20], Schaeffel’s Purkinje meter, pupil center for decentration and fixation axis for tilt). This shows that our measured data are similar to those published in other studies, so the dynamic Purkinje meter could be a suitable tool for determining the IOL position.

This evidence could be further strengthened by the presented case report of a 69-year-old patient. This subject was chosen to be presented in detail, as the IOL decentration values measured on the dynamic Purkinje meter were extreme in comparison to the values of the other examined subjects. The values of the IOL position were verified using CASIA2 and according to the real IOL center. The IOL tilt was measured practically the same using Purkinje meter (5.7°/240°) and using CASIA2 (5.7°/239°). On the other hand, the decentration magnitude was very different: 1.10 mm by Purkinje meter and 0.66 mm by CASIA2 (the difference in the decentration direction was only 10°); the IOL decentration according to the real IOL center was 0.94 mm (in direction 305°), and the difference from the Purkinje meter values was 0.16 mm (1°). The real correct value of IOL decentration is the one given according to the IOL center—therefore, the Purkinje meter provides more precise results than CASIA2.

The measured values of IOL position vary considerably across studies. The differences in the results could be caused by the variability of IOL types used in the studies and by the variability of the reference structures relative to which the position of the lens was determined. Overall, there are many published studies dealing with IOL tilt and decentration assessment using different methods; however, the comparison of these studies is questionable as no universal reference structure exists [5]. The reference structures used are mainly the pupil center or the pupillary axis [12,14,15,21,22,23,24], the visual axis [25] and the keratometric or topographic axis [9].

5. Conclusions

The correct position of the IOL and its stability in the capsular bag is crucial for the quality of the retinal image after cataract surgery. Although some IOL tilt and decentration are common, a higher extent of IOL tilt and decentration can significantly reduce a patient’s quality of vision, increase night vision symptoms, and lower the patient’s satisfaction. Due to the increasing demands of today’s society on visual quality and patient comfort, and due to the growing interest in the implantation of new and more complex IOL designs, determining the IOL position occupies an important position in current ophthalmological practice. The knowledge of the exact IOL position in the eye can help to explain and subsequently solve a refractive error, refine the IOL calculation before surgery on the second eye, study new IOL designs and optimize surgical procedures (centration and size of the capsulorhexis, placement of IOL haptics).

The upgraded dynamic Purkinje meter with a movable fixation target allows the assessment of a large range of IOL decentration and inclination. The device provides a sufficient, inexpensive alternative for IOL position determination to the commercially available systems and could serve as a good “helper” in cataract and refractive centers.