See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/309000816

Comparison of AL-Scan and IOLMaster 500 Partial Coherence Interferometry

Optical Biometers

Article  in  Journal of refractive surgery (Thorofare, N.J.: 1995) · October 2016

DOI: 10.3928/1081597X-20160712-03

CITATIONS READS

12 3,353

2 authors:

Kenneth J Hoffer Giacomo Savini

University of California, Los Angeles Fondazione G.B. Bietti
261 PUBLICATIONS   5,741 CITATIONS    250 PUBLICATIONS   7,219 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Protocols for Studies of Intraocular Lens Formula Accuracy View project

10,000 Eye study to test the Hoffer H-5 formula against the Barrett, Holladay 2, Haigis and Olsen formulas. View project

All content following this page was uploaded by Kenneth J Hoffer on 31 December 2017.

The user has requested enhancement of the downloaded file.

 ORIGINAL ARTICLE

Comparison of AL-Scan and IOLMaster 500

Partial Coherence Interferometry Optical
Biometers
Kenneth J. Hoffer, MD, FACS; Giacomo Savini, MD

ABSTRACT

PURPOSE: To investigate agreement between the ocu-

lar biometry measurements provided by a newer opti-
S ince 1999, optical biometry has become the standard
technique for axial length measurement.1-3 The main
reason for such popularity is not related to higher
accuracy, because its results are similar to those achieved by
immersion ultrasound biometry,4 but to the great advantage of
cal biometer, the AL-Scan (Nidek Co, Ltd., Gamagori,
Japan) and those provided by the IOLMaster 500 (Carl being a non-contact technique, which is associated with less
Zeiss Meditec, Jena Germany), which are both based on discomfort for the patient and less risk of corneal complica-
partial coherence interferometry. tions. Moreover, it is easier to perform for most surgeons and
technicians. Until 2009 the IOLMaster (Carl Zeiss Meditec,
METHODS: Axial length, corneal power, and anterior Jena, Germany), based on partial coherence interferometry,
chamber depth (corneal epithelium to lens) were mea- has been the only device measuring axial length by an optical
sured in 86 eyes of 86 patients scheduled for cataract
surgery using both biometers. All values were analyzed method. Since 2009, newer instruments have been introduced:
using a paired t test, the Pearson product moment cor- the LenStar (Haag-Streit AG, Köniz, Switzerland), the Aladdin
relation coefficient (r), and Bland–Altman plots. (Topcon EU, Visia Imaging, San Giovanni Valdarno, Italy),
the AL-Scan (Nidek Co. Ltd., Gamagori, Japan), the Galilei G6
RESULTS: The mean axial length values of both instru- (Ziemer, Port, Switzerland), the Argos (Movu, Santa Clara,
ments were exactly the same (23.46 ± 0.99 mm) for CA), the Pentacam AXL (Oculus, Wetzlar, Germany), and the
both) and showed excellent agreement and correlation.
On the contrary, the AL-Scan measured a steeper mean IOLMaster 700 (Carl Zeiss Meditec).
corneal power by 0.08 diopters (D) at the 2.4-mm zone Previous studies have shown that both the LenStar and
but by only 0.03 D at the 3.3-mm zone, only the former Aladdin provide similar measurements with respect to the
being statistically significant. The AL-Scan measured a IOLMaster 500,5-9 although slight differences in keratometry
deeper anterior chamber depth by 0.13 mm, which was (K) values, anterior chamber depth (ACD) (corneal epitheli-
statistically significant (P < .001).
um to lens), and axial length values require constant optimi-
CONCLUSIONS: Agreement between the two units was zation for each instrument when calculating intraocular lens
good. However, the small but statistically significant dif-
ference in corneal power (at the IOLMaster-comparable From Stein Eye Institute, University of California, Los Angeles, California
2.4-mm zone) and in the anterior chamber depth mea- (KJH); St. Mary’s Eye Center, Santa Monica, California (KJH); and G.B. Bietti
surement make lens constant optimization necessary Foundation IRCCS, Rome, Italy (GS).
when calculating the intraocular lens power by means
of theoretical formulas. Submitted: April 3, 2016; Accepted: June 23, 2016
Supported by the Italian Ministry of Health and Fondazione Roma (GS).
[J Refract Surg. 2016;32(10):694-698.]
Dr. Hoffer licenses the registered trademark name Hoffer® to Carl Zeiss-
Meditec (IOLMaster), Haag-Streit (LenStar), Oculus (Pentacam), Nidek
(AL-Scan), Tomey (OA-2000), Topcon EU Visia Imaging (Aladdin), Ziemer
(Galilei G6), and all A-scan biometer manufacturers to ensure proper pro-
gramming of the Hoffer Q and Hoffer H-5 formulas. Dr. Savini has no financial
or proprietary interest in the materials presented herein.
The authors thank H. John Shammas, MD, Lynwood, California, and Ralph
Khoury, MD, Saint Joseph University, Beirut, Lebanon, for patient data col-
lection.
Correspondence: Kenneth J. Hoffer, MD, FACS, St. Mary’s Eye Center, 411
Lincoln Blvd., Santa Monica, CA 90402. E-mail: [email protected]
doi:10.3928/1081597X-20160712-03

694 Copyright © SLACK Incorporated

 Partial Coherence Interferometry Optical Biometers/Hoffer & Savini

(IOL) power by theoretical formulas.10-12 To date, a few were recorded and any with a signal-to-noise ratio be-
studies investigated the Nidek optical biometer,13-15 low 2.0 were repeated. With respect to the keratometry,
but some of them suffer from methodological biases patients were requested to observe a yellow light and to
(in one case14 both eyes of most patients were enrolled blink to produce a continuous tear film, thus improving
and in another case15 IOL power calculation was not the reflectivity of the cornea. Six peripheral measuring
based on back-calculated constant optimization). points at a diameter of 2.3 mm were optimally focused
To validate the Nidek AL-Scan, in this prospective on the cornea as demonstrated by a green light from the
study, we aimed to confirm whether the biometric partial coherence interferometry “traffic light system.”
measurements have any statistically significant differ- Subsequent depression of the joystick button provided
ence in mean axial length, K values, and ACD. three consecutive K measurements and the mean of
these values was used for the IOL calculations. If any of
PATIENTS AND METHODS the 6 measurement points were not correctly identified,
PATIENTS the measurements were repeated.
This prospective observational study included con-
secutive adult patients scheduled for cataract surgery AL-SCAN MEASUREMENTS
between January and March 2014. Only one eye of each The technology of the Nidek AL-Scan (software ver-
patient was analyzed. The study was approved by the In- sion 1.03) is also based on partial coherence interferom-
stitutional Review Committee, all patients provided in- etry, with partial coherence superposition of light waves
formed consent, and the study complied with the tenets emitted from an 830-nm super luminescent diode laser,
of the Declaration of Helsinki. Exclusion criteria were to measure the axial length of the eye. It uses a 970-nm
previous corneal or intraocular surgery, any ocular sur- light-emitting diode (LED) for K assessment and a 525-
face disease (eg, dry eye) leading to irregular tear film, nm LED for determination of corneal diameter. Corneal
any corneal disease (eg, keratoconus or marginal pellu- power is measured by analyzing the images of double
cid degeneration), previous contact lens wear in the past mires of spots (360°) at diameters of 2.4 and 3.3 mm re-
month, and any retinal or optic nerve pathology affecting flected from the anterior surface of the cornea. In addi-
the postoperative visual acuity. Only eyes with a post- tion, the unit measures ACD, defined as the measurement
operative visual acuity of 20/40 or worse were enrolled. from the corneal epithelium to the anterior lens surface.
Each eye was evaluated on the same day using both Patients were carefully aligned for partial coherence
units. In each group, one unmasked examiner per- interferometry biometry measurements. The equip-
formed all measurements. For both instruments, the ment was optimally positioned as demonstrated by a
mean axial length, mean K, and ACD values were re- clear view of the anterior eye and the appearance of a
corded. All K values reported here are derived from quality control image, which indicated when the work-
the anterior corneal curvature using a 1.3375 kerato- ing distance of approximately 45 mm was achieved.
metric index of refraction. The patient was asked to fixate on the red fixation
lamp in the measuring window. This engaged align-
IOLMASTER 500 MEASUREMENTS ment software, which shows arrows to clearly indicate
The IOLMaster 500 (software version 5.2) optical bi- the direction in which the instrument must be moved
ometer uses partial coherence interferometry with a 780- to finetune the alignment. When the eye is aligned and
mm laser diode infrared light to measure axial length. in focus, a measurement starts automatically, taking a
The ACD is measured through a lateral slit-illumination measurement of consecutive six parameters: six axial
and is defined as the measurement from the corneal epi- length measurements, three K measurements, three
thelium to the anterior lens surface. The K readings are central corneal thickness and ACD measurements, one
calculated by analyzing the anterior corneal curvature pupil diameter measurement, and one corneal diam-
at six reference points in a hexagonal pattern at approxi- eter measurement, all under 10 seconds. The same pro-
mately the 2.3-mm optical zone. cedure was repeated three times and the average value
After ensuring the correct positioning of the patient for each parameter was used for statistical analysis.
against the chin and headrest, the partial coherence in-
terferometry was focused and coarsely aligned with the STATISTICAL ANALYSIS
participant’s eye using the overview mode. The patient Statistical analysis was performed using MedCalc
was directed to focus on the illuminated target. The axial software (version. 12.3.0.0; MedCalc Software, Ostend,
length measurement mode was activated and fine align- Belgium). All values were compared using a paired t
ment occurred while the patient was asked to observe test because all data followed a Gaussian distribution
the red fixation point. Five axial length measurements according to the Kolmogorov–Smirnov method. Cor-

Journal of Refractive SurgeryÊUÊ6œ°ÊÎÓ]Ê œ°Ê£ä]ÊÓä£È 695

Partial Coherence Interferometry Optical Biometers/Hoffer & Savini

TABLE 1
Comparison of Biometry Measures Between the AL-Scan and IOLMaster 500
Parameter AL-Scan Mean ± SD IOLMaster Mean ± SD Diff Pa 95% LoA CC r (P)
AL (mm) 23.46 ± 0.99 23.46 ± 0.99 +0.00 NS -0.03 to +0.03 0.9999 (< .001)
K 2.4 mm (D) 43.84 ± 1.49 43.76 ± 1.46 -0.08 .0152 -0.57 to +0.43 0.9851 (< .001)
K 3.3 mm (D) 43.79 ± 1.44 K@ 2.4 mm (43.76 ± 1.46) -0.03 NS -0.58 to + 0.53 0.9429 (< .001)
ACD (mm) 2.96 ± 0.38 2.83 ± 0.38 +0.13 .001 -0.44 to +0.18 0.9135 (< .001)
SD = standard deviation; Diff = difference; LoA = limits of agreement; CC = correlation coefficient; AL = axial length; NS = not statistically significant; K 2.4 =
corneal power at 2.4 mm ring for Nidek and IOLMaster; D = diopters; K 3.3 = corneal power at 3.3 mm ring for Nidek; ACD = anterior chamber depth (epithe-
lium to lens)
a
Paired t test.
The AL-Scan is manufactured by Nidek Co, Ltd., Gamagori, Japan, and the IOLMaster is manufactured by Carl Zeiss Meditec, Jena, Germany.

Figure 1. Bland–Altman plot for the axial length measurements. Figure 2. Bland–Altman plot for the 2.4-mm corneal power measure-
ments for the IOLMaster (Carl Zeiss Meditec, Jena, Germany) and
AL-Scan (Nidek Co, Ltd., Gamagori, Japan).
relation was used to quantify how well the measure-
ments by the two instruments varied. The Pearson
product moment correlation coefficient (r) was used ated. The mean age of the patients was 72 ± 9.1 years
to statistically evaluate each correlation. Agreement (range: 45 to 87 years). Table 1 describes the axial
was evaluated using the method of Bland and Alt- length, K values, and ACD measurements taken by the
man,16 who suggest plotting the differences between two instruments.
the measurements (y-axis) against their mean (x-axis). Axial length mean values were exactly the same for
The 95% limits of agreements (LoA) were defined as the two instruments (23.46 ± 0.99 mm), did not show
the mean ± 2 standard deviations (SDs) of the differ- any statistically significant difference, and showed ex-
ences between the two units. A P value less than .05 cellent agreement and correlation (r = 0.9999, P < .0001).
was considered statistically significant. Figure 1 shows the Bland–Altman plot for axial length.
Based on the result of a recent study,17 the SD of the The AL-Scan showed slightly steeper K measurements
differences in axial length, K, and ACD between two than the IOLMaster at the 2.4-mm zone by 0.08 D, which
optical biometers was 0.02 mm, 0.15 D, and 0.07 mm, was statistically significant. The difference was smaller
respectively. Using a two-sided level of significance (only 0.03 D) and not statistically significant when mea-
(!) at 0.05 and power (") at 90%, a sample size calcu- surements were taken at the 3.3-mm diameter. However,
lation indicates that a minimum of 48 patients would agreement was slightly better with measurements at the
be required to detect a mean difference of 0.01 mm, 2.4-mm diameter than those at the 3.3-mm diameter. Fig-
0.07 D, and 0.03 mm. ures 2-3 show the Bland–Altman plots for the 2.4- and
3.3-mm diameters, respectively. The AL-Scan showed
RESULTS deeper ACD measurements (mean difference: +0.13 ±
Eighty-six eyes of 86 patients (43 [50%] females, 47 0.43 mm), which was statistically significant (P = .0001).
[55%] right eyes) with a cataractous lens were evalu- Figure 4 shows the Bland–Altman plot for ACD.

696 Copyright © SLACK Incorporated

 Partial Coherence Interferometry Optical Biometers/Hoffer & Savini

Figure 3. Bland–Altman plot for the corneal power measurements for Figure 4. Bland–Altman plot for the anterior chamber depth measurements.
the AL-Scan (Nidek Co, Ltd., Gamagori, Japan) at 3.3 mm and IOLMaster
(Carl Zeiss Meditec, Jena, Germany) at 2.4 mm.
TABLE 2

DISCUSSION
Anterior Chamber Depth Measures
The current study shows that the two partial coher- Compared to IOLMaster
ence interferometry optical biometers provide similar Biometer Eyes Difference (mm) Status P
measurements, although agreement is not perfect other LenStar 61 +0.13 DEEPER .001
than for axial length. Some differences in K and ACD Aladdin 60 +0.16 DEEPER .001
do not allow us to consider the two instruments inter- AL-Scan 86 +0.13 DEEPER < .001
changeable.
The IOLMaster is manufactured by Carl Zeiss Meditec, Jena, Germany,
There was no statistically significant difference for the LenStar is manufactured by Haag-Streit, the Aladdin is manufactured
mean axial length. Agreement was excellent as the by Topcon EU Visia Imaging, Capelle aan den Ijssel, Netherlands, and the
AL-Scan is manufactured by Nidek Co, Ltd., Gamagori, Japan.
95% LoA showed that the difference would be low-
er than 0.2 mm in 95% of cases. In this regard, our
data confirm the results of previous studies compar- ter (approximately 2.3 mm) than other instruments, such
ing these two devices.13-15 Our data are also similar to as the Lenstar or the Pentacam (which analyze a diameter
those previously reported between the IOLMaster 500 of approximately 3 mm). However, the difference in the
and the LenStar5-7 and between the IOLMaster 500 and current comparison cannot be related to the analyzed
the Aladdin.8,9 diameter, which is almost the same for the two partial
Some differences were found for the mean K val- coherence interferometry biometers. Therefore, the dif-
ues, which were steeper with the AL-Scan in both opti- ference is likely to depend on the method used to analyze
cal zones. Although agreement was good and the mean the mires of spots reflected from the cornea.
difference may not seem clinically significant, actually As regards ACD, the AL-Scan provided deeper mean
this difference in the 2.4-mm diameter is sufficient to re- values compared to the IOLMaster 500. This is likely
quire constant optimization to nullify systematic differ- to depend on the different technology used to measure
ences between the two devices. Our results are in good the ACD; the IOLMaster measures ACD through a lateral
accordance with previous studies, where the AL-Scan slit illumination, whereas the AL-Scan measures ACD
was found to provide steeper K measurements than the using Scheimpflug imaging. Similarly, both the LenStar
IOLMaster.13-15 On the other hand, in two of three stud- and, to a lesser extent, the Aladdin gave mean deeper
ies those differences were not statistically significant,13,15 ACD values than the IOLMaster 50059 (Table 2). Deeper
whereas we found the difference at the 2.4-mm diameter ACD measurements by the AL-Scan with respect to the
to be statistically significant. It is hard to explain why the IOLMaster had already been reported by Huang et al.13
AL-Scan provides steeper K measurements than the IOL- and Kaswin et al., 15 whereas the opposite result had
Master. In most published studies, the IOLMaster provid- been found by Srivannaboon et al.14
ed mean steeper K values than other technologies, such This study is limited by the fact that we did not
as Scheimpflug imaging and Placido-disk corneal topog- evaluate the accuracy of astigmatic axis measurement
raphy.5,18,19 These differences had been related to the fact or the accuracy of the two devices in IOL power cal-
that the IOLMaster 500 analyzes a smaller corneal diame- culation, but these will be the subjects of further in-

Journal of Refractive SurgeryÊUÊ6œ°ÊÎÓ]Ê œ°Ê£ä]ÊÓä£È 697

Partial Coherence Interferometry Optical Biometers/Hoffer & Savini

vestigation. Also, our sample did not include myopic 7. Buckhurst PJ, Wolffsohn JS, Shah S, Naroo SA, Davies LN,
Berrow EJ. A new optical low coherence reflectometry de-
eyes with an axial length greater than 27.50 mm, so our vice for ocular biometry in cataract patients. Br J Ophthalmol.
results cannot necessarily be applied to those cases. 2009;93:949-953.
Finally, we did not investigate some of the potentially 8. Hoffer KJ, Shammas HJ, Savini G. Comparison of a new optical
advantageous features of the AL-scan such as the at- biometer and partial coherence interferometry: a multicenter
tached ultrasound probe for eyes that cannot be mea- study of Caucasian and Chinese subjects. J Cataract Refract
Surg. 2016;42:62-67.
sured optically.
9. Mandal P, Berrow EJ, Naroo SA, et al. Validity and repeatability
Both optical biometers provide close mean measure- of the Aladdin ocular biometer. Br J Ophthalmol. 2014;98:256-
ments of the main biometric parameters of concern in 258. Erratum in: Br J Ophthalmol. 2015;99:1746.
IOL power calculation. Although there is no difference 10. Hoffer KJ. The Hoffer Q formula: a comparison of theoretic and
in axial length measurement, the small differences in regression formulas. J Cataract Refract Surg. 1993;19:700-712.
Errata in: J Cataract Refract Surg. 1994;20:677 and 2007;33:2-3.
K values and ACD should not be overlooked and war-
rant constant optimization for IOL power calculation. 11. Holladay JT, Praeger TC, Chandler TY, Musgrove KH, Lewis
JW, Ruiz RS. A three-part system for refining intraocular lens
power calculations. J Cataract Refract Surg. 1988;14:17-24.
AUTHOR CONTRIBUTIONS 12. Retzlaff JA, Sanders DR, Kraff MC. Development of the SRK/T
Study concept and design (KJH); analysis and interpretation of intraocular lens implant power calculation formula. J Cataract
data (KJH, GS); writing the manuscript (KJH); critical revision of the Refract Surg. 1990;16:528.
manuscript (KJH, GS); administrative, technical, or material support 13. Huang J, Savini G, Li J, et al. Evaluation of a new optical biometry
device for measurements of ocular components, and its compari-
(KJH); supervision (KJH)
son with IOLMaster. Br J Ophthalmol. 2014;98:1277-1281.
14. Srivannaboon S, Chirapapaisan C, Chonpimai P, Koodkaew S.
REFERENCES Comparison of ocular biometry and intraocular lens power us-
1. Haigis W. Optical biometry using partial coherence interferom- ing a new biometer and a standard biometer. J Cataract Refract
etry. In: Shammas HJ, ed. Intraocular Lens Power Calculations. Surg. 2014;40:709-715.
Thorofare, NJ: SLACK Incorporated; 2004:141-157.
15. Kaswin G, Rousseau A, Mgarrech M, Barreau E, Labetoulle M.
2. Packer M, Fine IH, Hoffman RS, Coffman PG, Brown LK. Im- Biometry and intraocular lens power calculation results with a
mersion A-scan compared with partial coherence interferom- new optical biometry device: comparison with the gold stan-
etry; outcomes analysis. J Cataract Refract Surg. 2002;28:239- dard. J Cataract Refract Surg. 2014;40:593-600.
242.
16. Bland JM, Altman DG. Statistical methods for assessing agree-
3. Bhatt AB, Schefler AC, Feuer WJ, Yoo SH, Murray TG. Com- ment between two methods of clinical measurement. Lancet.
parison of predictions made by the Intraocular Lens Master and 1986;1:307–310.
ultrasound biometry. Arch Ophthalmol. 2008;126:929-933.
17. Huang J, McAlinden C, Su B, et al. The effect of cycloplegia
4. Haigis W, Lege B, Miller N, Schneider B. Comparison of im- on the Lenstar and the IOLMaster biometry. Optom Vis Sci.
mersion ultrasound biometry and partial coherence interferom- 2012;89:1691-1696.
etry for intraocular lens power calculation according to Haigis.
Graefes Arch Clin Exp Ophthalmol. 2000;238:765-773. 18. Shirayama M, Wang L, Weikert MP, Koch DD. Comparison of
corneal powers obtained by 4 different devices. Am J Ophthal-
5. Hoffer KJ, Shammas HJ, Savini G. Comparison of 2 instruments mol. 2009;148:528-535.
for measuring axial length. J Cataract Refract Surg. 2010;36:644-
664. Erratum in: J Cataract Refract Surg. 2010;36:1066. 19. Savini G, Barboni P, Carbonelli M, Hoffer KJ. Accuracy of
Scheimpflug corneal power measurements for intraocular lens
6. Holzer MP, Mamusa M, Auffarth GU. Accuracy of a new partial power calculation. J Cataract Refract Surg. 2009;35:1193-1197.
coherence interferometry analyser for biometric measurements. Erratum in: J Cataract Refract Surg. 2009;35:1483.
Br J Ophthalmol. 2009;93:807-810.

698 Copyright © SLACK Incorporated

View publication stats