- Email: [email protected]
Repeatability and reproducibility of optical biometry implemented in a new optical coherence tomographer and comparison with a optical low-coherence reflectometer
1619
ARTICLE
Repeatability and reproducibility of optical biometry implemented in a new optical coherence tomographer and comparison with a optical low-coherence reflectometer Piotr Kanclerz, MD, PhD, Kenneth J. Hoffer, MD, Jos J. Rozema, MSc, PhD, Katarzyna Przewłocka, BSc, Giacomo Savini, MD
Purpose: To examine the repeatability and reproducibility of biometry provided by a new anterior–posterior segment spectraldomain optical coherence tomography (SD-OCT) device (Revo NX) and compare it with that provided by a validated optical lowcoherence reflectometry (OLCR) biometer (Lenstar LS 900). Setting: Hygeia Clinic, Gdansk, Poland. Design: Prospective evaluation of diagnostic test. Methods: The axial length (AL), lens thickness (LT), central corneal thickness (CCT), and anterior chamber depth (ACD) (corneal epithelium to lens) were measured in eyes of volunteers. Two skilled operators obtained 3 measurements in random order with the SD-OCT device. Repeatability was calculated for each observer, while reproducibility was assessed by a change in observer. To compare agreement between the SD-OCT device and OLCR device, the first observer took 3 measurements with both devices.
ince the first optical biometer (IOLMaster, Carl Zeiss Meditec AG) was introduced in 1999, several similar instruments have been developed. The first devices combined optical biometry (to measure axial distances) and automated keratometry (to measure corneal curvature and calculate corneal power). These include the Lenstar LS 900 (Haag-Streit AG)1 and AL-Scan (Nidek Co., Ltd).2 Subsequently, optical biometry was combined with Placido-disk corneal topography on 2 devices, the Aladdin (Europe Medical B.V.)3 and OA-2000 (Tomey Corp.),4 and with a rotating Scheimpflug camera on 2 other instruments, the Galilei G6 (Ziemer Ophthalmic Systems AG)5 and Pentacam AXL (Oculus Optikger€ate GmbH).6 More
S
Results: The study evaluated 63 eyes of 63 volunteers. The mean AL measurement (C0.11 mm G 0.02 [SD]) and mean ACD measurement (0.05 G 0.04 mm) were significantly higher with the SD-OCT device than with the OLCR device; the CCT measurements did not differ between the devices. A very strong correlation was found for all parameters, while Bland-Altman analysis showed narrow 95% limits of agreement for AL (0.06 to 0.15 mm), ACD (0.02 to 0.12 mm), LT (0.23 to 0.17 mm), and CCT (7.83 to 6.17 mm). Conclusions: The new SD-OCT device provided repeatable and reproducible AL, ACD, LT, and CCT measurements. The results with the SD-OCT device showed strong correlations with those obtained with the OLCR device. However, the AL and ACD measurements of the 2 devices cannot be considered interchangeable. J Cataract Refract Surg 2019; 45:1619–1624 Q 2019 ASCRS and ESCRS
recently, a new generation of optical biometers was developed based on optical coherence tomography (OCT); these include the IOLMaster 700 (Carl Zeiss Meditec AG),7 the Argos (Movu, Inc.),8 and the Anterion (Heidelberg Engineering GmbH). Although the latter group provides excellent anterior segment OCT images, none is able to generate high-definition scans of the retina and optic nerve and give, at the same time, the biometric parameters required to calculate intraocular lens (IOL) power. The recently introduced Revo NX device (Optopol Technology S.A.) combines anterior and posterior segment spectraldomain OCT (SD-OCT) with optical biometry.9 Having one instrument to perform both tasks would be an
Submitted: April 3, 2019 | Final revision submitted: June 21, 2019 | Accepted: July 1, 2019 cka), Gdan sk, Poland; the Stein Eye Institute (Hoffer), University of California, Los Angeles, and St. Mary’s Eye Center (Hoffer), From Hygeia Clinic (Kanclerz, Przew1o Santa Monica, California, USA; the Department of Ophthalmology (Rozema), Antwerp University Hospital, Edegem, and Faculty of Medicine and Health Sciences (Rozema), University of Antwerp, Belgium; the IRCCS Fondazione Bietti (Savini), Rome, Italy. The contribution of IRCCS GB Bietti Foundation was supported by the Italian Ministry of Health and Fondazione Roma, Italy. sk, Poland. Email: [email protected]. Corresponding author: Piotr Kanclerz, MD, PhD, ul. Jaskowa Dolina 57, 80-286 Gdan Q 2019 ASCRS and ESCRS Published by Elsevier Inc.
0886-3350/$ - see frontmatter https://doi.org/10.1016/j.jcrs.2019.07.002
1620
OCT VS OLCR
Figure 1. Optical biometry results obtained with the SD-OCT device in a healthy patient with a clear lens (AL Z axial length; ACD Z anterior chamber depth; LT Z lens thickness; CCT Z central corneal thickness).
advantage to clinicians because they would have to purchase only one device rather than two. The purpose of this study was to examine the repeatability and reproducibility of the measurements provided by the new Revo NX optical biometer and compare them with those of a validated optical biometer,10 the Lenstar LS 900, which is based on optical low-coherence reflectometry (OLCR). PARTICIPANTS AND METHODS This prospective study enrolled volunteers with healthy eyes at Hygeia Clinic, Gda nsk, Poland, between July 2018 and August 2018. The study adhered to the tenets of the Declaration of Helsinki for the use of human participants in biomedical research and was approved by the local research ethics board. All participants signed informed consent forms after receiving an explanation of the purpose of the study. Exclusion criteria were any ocular disease (including cataract), corrected distance visual acuity worse than 20/25, previous ocular surgery, and trauma. Before being enrolled, all eyes had a complete ophthalmologic examination consisting of subjective refraction, noncontact tonometry, slitlamp evaluation, and ophthalmoscopy. Instruments The Revo NX SD-OCT device has an axial resolution of 5 mm, transverse resolution of 12 mm, and single scan depth of 2.4 mm
and obtains 110 000 scans per second. A superluminescent laser diode (830 nm) is used as the light source. The optical biometry program within the device measures the axial length (AL), anterior chamber depth (ACD; measured from the epithelium to lens), lens thickness (LT), and central corneal thickness (CCT). For each measurement, it performs 10 B-scans to calculate a precise average value (Figure 1). The Lenstar LS-900 optical low-coherence reflectometry (OLCR) biometer is a superluminescent diode laser (820 nm) that measures the AL, LT, CCT, and aqueous depth. The device also provides keratometry values, pupil size, and corneal diameter measurements. To allow comparison of the ACD values determined by the 2 devices, the ACD for the OLCR device was calculated as the sum of aqueous depth and CCT.11
Measurement Technique Measurements were performed with the 2 devices in random order. Only right eyes were analyzed, and all scans were taken between 15:00 and 19:00. Two skilled operators, in random order, performed 3 scans with the SD-OCT device to obtain 3 measurements. Repeatability was calculated for each observer, while reproducibility was assessed by a change in observer. To assess the agreement between the SD-OCT device and the OLCR device, the first observer took 3 measurements consecutively with both devices.
Table 1. Repeatability outcomes for 3 biometry measurements by each of 2 observers obtained using the spectral-domain optical coherence tomography device. Parameter AL (mm) Observer 1 Observer 2 ACD (mm) Observer 1 Observer 2 LT (mm) Observer 1 Observer 2 CCT (mm) Observer 1 Observer 2
Mean ± SD (mm)
Sr
r
CoV (%)
ICC
23.37 G 1.07 23.37 G 1.06
0.01 0.02
0.04 0.04
0.06 0.07
0.986 0.985
3.40 G 0.36 3.41 G 0.36
0.01 0.02
0.04 0.05
0.43 0.49
0.961 0.956
4.09 G 0.37 4.09 G 0.38
0.03 0.03
0.08 0.07
0.67 0.62
0.931 0.938
550.51 G 30.51 549.51 G 31.39
2.58 2.68
7.16 7.42
0.47 0.49
0.922 0.920
ACD Z anterior chamber depth; AL Z axial length; CCT Z central corneal thickness; CoV Z coefficient of variation; ICC Z intraclass correlation coefficients; LT Z lens thickness; r Z repeatability limit; Sr Z repeatability
Volume 45 Issue 11 November 2019
1621
OCT VS OLCR
Table 2. Reproducibility outcomes for biometry measurements obtained using the spectral-domain optical coherence tomography device. Parameter
Mean Difference ± SD
SR
R
CoV (%)
ICC
AL (mm) ACD (mm) LT (mm) CCT (mm)
0.00 G 0.02 0.01 G 0.03 0.00 G 0.04 1.00 G 5.43
0.01 0.01 0.01 1.76
0.02 0.04 0.04 4.88
0.03 0.38 0.37 0.32
0.993 0.965 0.962 0.945
ACD Z anterior chamber depth; AL Z axial length; CCT Z central corneal thickness; CoV Z coefficient of variation; ICC Z intraclass correlation coefficients; LT Z lens thickness; R Z reproducibility limit; Sr Z reproducibility
Statistical Analysis Statistical analysis was performed using Prism software (GraphPad Software, Inc.) and MedCalc software (MedCalc Software bvba). The results are presented as the mean G SD. The repeatability, repeatability limit, reproducibility, and reproducibility limit were calculated as recommended by McAlinden et al.12 The coefficient of variation (CoV) was determined as the ratio of the repeatability/reproducibility to the mean value (lower CoV stands for better reliability).13 The intraclass correlation coefficient (ICC) was defined as the ratio of variance between subjects to the sum of the pooled intraparticipant variance and the interparticipant variance. An ICC less than 0.75 indicates poor agreement; an ICC from 0.75 to 0.90 is considered as moderate, and an ICC of 0.90 or above is considered high.14 Comparison between the devices was performed using correlation coefficients, the Bland-Altman method, and paired t tests given that all parameters were normally distributed in the Kolmogorov-Smirnov test. A P value less than .05 was considered statistically significant.
RESULTS The study enrolled 65 participants. Two participants were excluded from the study; no measure for LT could be obtained in 1 eye and 1 participant withdrew consent for participation. Therefore, the results of 63 participants (44 women) were analyzed. The mean age was 41.8 G 13.8 years. The mean spherical equivalent refraction was 0.11 G 2.02 diopters (D) (range 5.38 to C4.63 D). The AL, ACD, LT, and CCT measurements obtained with SD-OCT device had high repeatability for both observers (Table 1). The CoV was less than 0.7% for all
parameters, while the ICC was higher than 0.92 for all parameters. Table 2 shows the reproducibility of the outcomes measured by the SD-OCT device. The reproducibility limit values were excellent, showing low variability. There was a statistically significant difference in AL, ACD, and LT measurements between the SD-OCT device and OLCR device (P ! .01). The mean AL and ACD measurements were higher with the SD-OCT device (Table 3). The CCT measurements were not significantly different between the devices (P Z .07). The correlation between the systems was very strong for AL (r Z 0.9997), ACD (r Z 0.9953), LT (r Z 0.9626), and CCT (r Z 0.9922). Bland-Altman analysis showed narrow 95% limits of agreement for AL, ACD, LT, and CCT measurements. Figures 2 to 5 show the Bland-Altman plots for these parameters. DISCUSSION Information about the repeatability and reproducibility of measurements and agreement with validated devices is essential when a new instrument becomes commercially available. Our data show that the measurements provided by the new Revo NX SD-OCT optical biometer offer high repeatability and reproducibility and a very strong correlation with those by the Lenstar LS 900 OLCR device. Although the agreement was good for all parameters except AL, a statistically significant difference was detected for all
Table 3. Mean values, mean differences, 95% LoA, and correlation for differences between the SD-OCT device and the OLCR device. Parameter/Device AL (mm) SD-OCT OLCR ACD (mm) SD-OCT OLCR LT (mm) SD-OCT OLCR CCT (mm) SD-OCT OLCR
Mean ± SD
Mean Difference ± SD
95% LoA
r Value
P Value
23.38 G 1.07 23.27 G 1.07
0.11 G 0.02
0.06, 0.15
0.9997
!.00001
3.39 G 0.36 3.34 G 0.36
0.05 G 0.04
0.02, 0.12
0.9953
!.00001
4.09 G 0.37 4.12 G 0.37
0.03 G 0.10
0.23, 0.17
0.9626
!.00001
550.51 G 30.51 551.34 G 30.50
0.83 G 3.57
7.83, 6.17
0.9922
!.00001
ACD Z anterior chamber depth; AL Z axial length; CCT Z central corneal thickness; LoA Z limits of agreement; LT Z lens thickness; OLCR Z optical lowcoherence reflectometry; r Z correlation coefficient; SD-OCT Z spectral-domain optical coherence tomography
Volume 45 Issue 11 November 2019
1622
OCT VS OLCR
Figure 2. Bland-Altman plot of agreement in axial length measurements between SD-OCT and OLCR devices (OLCR Z optical low-coherence reflectometry device; SD-OCT Z spectral-domain optical coherence tomography device).
Figure 3. Bland-Altman plot of agreement in anterior chamber depth measurements between SD-OCT and OLCR devices (OLCR Z optical low-coherence reflectometry device; SDOCT Z spectral-domain optical coherence tomography device).
of the 4 parameters evaluated in this study (AL, ACD, CCT LT). The largest dissimilarities were observed for the AL (mean difference C0.11 G 0.02 mm) and ACD (mean difference C0.05 G 0.04 mm). Such differences have not only statistical significance but also clinical significance and should not be underestimated; a shift of 0.1 mm in AL measurement transposes to approximately a 0.27 D change in the spectacle plane (for normal eye dimensions).15 Variations in biometric parameters have been reported with many optical biometers.16–25 In a study by Goebels et al.,21 AL values obtained with partial coherence interferometry (PCI) were on average 0.07 mm higher than those obtained with OLCR. Ortiz et al.18 reported that the AL determined with PCI was on average 0.04 mm shorter than that obtained with OLCR biometry. In a meta-analysis by Rozema et al.,10 the Lenstar OLCR device measurements were on average 0.02 G 0.01 mm (SD) higher than the IOLMaster PCI measurements, which was significantly equal. Table 4 shows the results in studies reporting differences in AL and ACD optical biometry measurements.1–3,7,16,18,21–25 The differences between the devices in our study could presumably be associated with the use of a dissimilar group refractive index to convert the optical path length into axial distances.26 Other potential reasons include the minimally different optical wavelength used by the devices or a dissimilar technique for automatic structure
border detection. These issues are likely a matter of calibration and should play no major role in the usefulness of the device. The difference in AL and ACD values in our study requires specific constant optimization for IOL power theoretical formulas that include them as predictors of the IOL position. This study and the SD-OCT optical biometer have limitations. First, we did not evaluate patients with eye disease, including cataract; therefore, our findings can be applied only to healthy eyes. Second, keratometric measurements and IOL calculation were not available in the device at the time of the study. Nevertheless, we believe this feature is an interesting development of OCT because most clinics likely have a validated keratometry device and theoretically could use the obtained data in IOL calculation formulas. Future software versions of this device enabling keratometric measurements will require further validation. In conclusion, the new Revo NX SD-OCT optical biometer provided repeatable and reproducible AL, ACD, LT, and CCT measurements. The results obtained with the SD-OCT biometer had a very strong correlation with those obtained with the Lenstar OLCR device; however, AL measurements, and to a lesser extent ACD measurements, cannot be considered interchangeable.
Figure 4. Bland-Altman plot of agreement in lens thickness measurements between SD-OCT and OLCR devices (OLCR Z optical low-coherence reflectometry device; SD-OCT Z spectral-domain optical coherence tomography device).
Figure 5. Bland-Altman plot of agreement in central corneal thickness measurements between SD-OCT and OLCR devices (OLCR Z optical low-coherence reflectometry device; SDOCT Z spectral-domain optical coherence tomography device).
Volume 45 Issue 11 November 2019
1623
OCT VS OLCR
Table 4. Studies reporting differences in AL and ACD measurements by optical biometry devices. Differences are presented as IOLMaster minus other device (mm). The IOLMaster 500 was used for all except 1 study,7 in which the 700 version was used. Axial Length (mm) Study*/Year Buckhurst 200923 Hoffer 20101 Rabsilber 201025 Epitropoulos 201416 Goebels 201521 Goebels 201521 Hoffer 20167 Hoffer 20162 Hoffer 20163 Whang 201822 Ortiz 201918 Ortiz 201918 Muzyka-Wozniak 201924
Anterior Chamber Depth (mm)
Device
n
Mean Diff
95% LoA
Mean Diff
Lenstar LS 900 Lenstar LS 900 Lenstar LS 900 Lenstar LS 900 Lenstar LS 900 OA-2000 Lenstar LS 900 AL-Scan Aladdin Argos Lenstar LS 900 Aladdin Pentacam AXL
112 50 100 105 138 138 183 86 60† 153 197 197 97
0.01 G 0.02 0.02 0.0 G 0.08z 0.03 G 0.03 0.02 G 0.03 0.05 G 0.09 0.01 0.0z 0.01z 0.03 0.04 G 0.26 0.04 G 0.13 0.01 G 0.04
d 0.03, 0.01 d 0.08, 0.03 d d 0.06, 0.03 0.03, 0.03 0.12, 0.09 d d d 0.01, 0.02
0.1 G 0.02 0.15 0.05 G 0.11 0.1 G 0.18 0.1 G 0.17 0.18 G 0.21 0.03 0.13 0.16 G 0.29 0.01z 0.13 G 0.23 0.1z 0.01 G 0.1
95% LoA d 0.21, 0.08 d 0.45, 0.25 d 0.09, 0.04 0.18, 0.44 0.78, 0.42 d d d 0.04, 0.02
Means G SD Diff Z difference; LoA Z limits of agreement *First Author † States cohort z Not statistically significant
WHAT WAS KNOWN Optical methods are considered to be the gold standard for preoperative biometry. At present, it is necessary to buy a separate optical coherence tomographer (OCT) device for anterior and posterior segment imaging and another for optical biometry.
WHAT THIS PAPER ADDS Optical biometry can be used in a new anterior and posterior segment spectral-domain (SD) OCT device. The new SD-OCT biometer provided repeatable and reproducible measurements of axial length, anterior chamber depth, lens thickness, and central corneal thickness. The results obtained with the new biometer showed a very strong correlation with a validated optical biometer. However, they cannot be considered interchangeable.
REFERENCES 1. Hoffer KJ, Shammas HJ, Savini G. Comparison of 2 laser instruments for measuring axial length. J Cataract Refract Surg 2010; 36:644–648; erratum, 1066 2. Hoffer KJ, Savini G. Comparison of AL-Scan and IOLMaster 500 partial coherence interferometry optical biometers. J Refract Surg 2016; 32:694–698 3. Hoffer KJ, Shammas HJ, Savini G, Huang J. Multicenter study of optical low-coherence interferometry and partial-coherence interferometry optical biometers with patients from the United States and China. J Cataract Refract Surg 2016; 42:62–67 4. Huang J, Savini G, Hoffer KJ, Chen H, Lu W, Hu Q, Bao F, Wang Q. Repeatability and interobserver reproducibility of a new optical biometer based on swept-source optical coherence tomography and comparison with IOLMaster. Br J Ophthalmol 2017; 101:493–498 5. Savini G, Negishi K, Hoffer KJ, Schiano Lomoriello D. Refractive outcomes of intraocular lens power calculation using different corneal power measurements with a new optical biometer. J Cataract Refract Surg 2018; 44:701–708
€ller M, Kohnen T. Com6. Shajari M, Cremonese C, Petermann K, Singh P, Mu parison of axial length, corneal and anterior chamber depth measurements of two recently introduced devices to a known biometer. Am J Ophthalmol 2017; 178:58–64 7. Hoffer KJ, Hoffmann PC, Savini G. Comparison of a new optical biometer using swept-source optical coherence tomography and a biometer using optical low-coherence reflectometry. J Cataract Refract Surg 2016; 42:1165–1172 8. Shammas HJ, Ortiz S, Shammas MC, Kim SH, Chong C. Biometry measurements using a new large-coherence–length swept-source optical coherence tomographer. J Cataract Refract Surg 2016; 42:50–61 9. Kanclerz P. Optical biometry in a commercially available anterior and posterior segment optical coherence tomography (OCT) device. Clin Exp Optom 2019 [Epub ahead of print] 10. Rozema JJ, Wouters K, Mathysen DGP, Tassignon M-J. Overview of the repeatability, reproducibility, and agreement of the biometry values provided by various ophthalmic devices. Am J Ophthalmol 2014; 158:1111–1120 11. Hoffer KJ. Definition of ACD [letter]. Ophthalmology 2011; 118:1484 12. McAlinden C, Khadka J, Pesudovs K. Precision (repeatability and reproducibility) studies and sample-size calculation [guest editorial]. J Cataract Refract Surg 2015; 41:2598–2604 13. Bland JM, Altman DG. Measurement error [statistics notes]. BMJ 1996; 312:1654 €ller R, Bu €ttner P. A critical discussion of intraclass correlation coeffi14. Mu cients. Stat Med 1994; 13:2465–2476 15. Olsen T. Calculation of intraocular lens power: a review. Acta Ophthalmol Scand 2007; 85:472–485 16. Epitropoulos A. Axial length measurement acquisition rates of two optical biometers in cataractous eyes. Clin Ophthalmol 2014; 8:1369–1376 17. Holzer MP, Mamusa M, Auffarth GU. Accuracy of a new partial coherence interferometry analyser for biometric measurements. Br J Ophthalmol 2009; 93:807–810 ~a V, Corrales MI, Ochoa M, Rodriguez CJ. 18. Ortiz A, Galvis V, Tello A, Vian Comparison of three optical biometers: IOLMaster 500, Lenstar LS 900 and Aladdin. Int Ophthalmol 2019; 39:1809–1818 19. Kongsap P. Comparison of a new optical biometer and a standard biometer in cataract patients. Eye Vis 2016; 3:27 20. Savini G, Hoffer KJ, Schiano-Lomoriello D. Agreement between lens thickness measurements by ultrasound immersion biometry and optical biometry. J Cataract Refract Surg 2018; 44:1463–1468 €ller M, Eppig T, Cayless A, Seitz B, Langenbucher A. 21. Goebels S, Pattmo Comparison of 3 biometry devices in cataract patients. J Cataract Refract Surg 2015; 41:2387–2393
Volume 45 Issue 11 November 2019
1624
OCT VS OLCR
22. Whang W-J, Yoo Y-S, Kang M-J, Joo C-K. Predictive accuracy of partial coherence interferometry and swept-source optical coherence tomography for intraocular lens power calculation. Sci Rep 2018; 8:13732 23. Buckhurst PJ, Wolffsohn JS, Shah S, Naroo SA, Davies LN, Berrow EJ. A new optical low coherence reflectometry device for ocular biometry in cataract patients. Br J Ophthalmol 2009; 93:949–953 24. Muzyka-Wozniak M, Oleszko A. Comparison of anterior segment parameters and axial length measurements performed on a Scheimpflug device with biometry function and a reference optical biometer. Int Ophthalmol 2019; 39:1115–1122 25. Rabsilber TM, Jepsen C, Auffarth GU, Holzer MP. Intraocular lens power calculation: clinical comparison of 2 optical biometry devices. J Cataract Refract Surg 2010; 36:230–234 26. Haigis W, Lege B, Miller N, Schneider B. Comparison of immersion ultrasound biometry and partial coherence interferometry for intraocular lens calculation according to Haigis. Graefes Arch Clin Exp Ophthalmol 2000; 238:765–773
Disclosures: Dr. Kanclerz receives non-financial support from Visim and Optopol Technology SA. To ensure accurate programming of his formulas, Dr. Hoffer licenses the registered trademark
Volume 45 Issue 11 November 2019
name HofferÒ to Carl Zeiss Meditec AG (IOLMasters), Haag-Streit AG (LenStar/EyeStar), Heidelberg Engineering, Inc. (Anterion), €te Movu, Inc. (Argos), Nidek, Inc. (AL-Scan), Oculus Optikgera GmbH (Pentacam AXL), Tomey Corp. (OA-2000), Topcon Europe Medical B.V./Visia Imaging S.r.l. (Aladdin), Ziemer Ophthalmic Systems AG (Galilei G6), and all A-scan biometer manufacturers. None of the other authors has a financial or proprietary interest in any material or method mentioned.
First author: Piotr Kanclerz, MD, PhD sk, Poland Hygeia Clinic , Gdan
ARTICLE
Repeatability and reproducibility of optical biometry implemented in a new optical coherence tomographer and comparison with a optical low-coherence reflectometer Piotr Kanclerz, MD, PhD, Kenneth J. Hoffer, MD, Jos J. Rozema, MSc, PhD, Katarzyna Przewłocka, BSc, Giacomo Savini, MD
Purpose: To examine the repeatability and reproducibility of biometry provided by a new anterior–posterior segment spectraldomain optical coherence tomography (SD-OCT) device (Revo NX) and compare it with that provided by a validated optical lowcoherence reflectometry (OLCR) biometer (Lenstar LS 900). Setting: Hygeia Clinic, Gdansk, Poland. Design: Prospective evaluation of diagnostic test. Methods: The axial length (AL), lens thickness (LT), central corneal thickness (CCT), and anterior chamber depth (ACD) (corneal epithelium to lens) were measured in eyes of volunteers. Two skilled operators obtained 3 measurements in random order with the SD-OCT device. Repeatability was calculated for each observer, while reproducibility was assessed by a change in observer. To compare agreement between the SD-OCT device and OLCR device, the first observer took 3 measurements with both devices.
ince the first optical biometer (IOLMaster, Carl Zeiss Meditec AG) was introduced in 1999, several similar instruments have been developed. The first devices combined optical biometry (to measure axial distances) and automated keratometry (to measure corneal curvature and calculate corneal power). These include the Lenstar LS 900 (Haag-Streit AG)1 and AL-Scan (Nidek Co., Ltd).2 Subsequently, optical biometry was combined with Placido-disk corneal topography on 2 devices, the Aladdin (Europe Medical B.V.)3 and OA-2000 (Tomey Corp.),4 and with a rotating Scheimpflug camera on 2 other instruments, the Galilei G6 (Ziemer Ophthalmic Systems AG)5 and Pentacam AXL (Oculus Optikger€ate GmbH).6 More
S
Results: The study evaluated 63 eyes of 63 volunteers. The mean AL measurement (C0.11 mm G 0.02 [SD]) and mean ACD measurement (0.05 G 0.04 mm) were significantly higher with the SD-OCT device than with the OLCR device; the CCT measurements did not differ between the devices. A very strong correlation was found for all parameters, while Bland-Altman analysis showed narrow 95% limits of agreement for AL (0.06 to 0.15 mm), ACD (0.02 to 0.12 mm), LT (0.23 to 0.17 mm), and CCT (7.83 to 6.17 mm). Conclusions: The new SD-OCT device provided repeatable and reproducible AL, ACD, LT, and CCT measurements. The results with the SD-OCT device showed strong correlations with those obtained with the OLCR device. However, the AL and ACD measurements of the 2 devices cannot be considered interchangeable. J Cataract Refract Surg 2019; 45:1619–1624 Q 2019 ASCRS and ESCRS
recently, a new generation of optical biometers was developed based on optical coherence tomography (OCT); these include the IOLMaster 700 (Carl Zeiss Meditec AG),7 the Argos (Movu, Inc.),8 and the Anterion (Heidelberg Engineering GmbH). Although the latter group provides excellent anterior segment OCT images, none is able to generate high-definition scans of the retina and optic nerve and give, at the same time, the biometric parameters required to calculate intraocular lens (IOL) power. The recently introduced Revo NX device (Optopol Technology S.A.) combines anterior and posterior segment spectraldomain OCT (SD-OCT) with optical biometry.9 Having one instrument to perform both tasks would be an
Submitted: April 3, 2019 | Final revision submitted: June 21, 2019 | Accepted: July 1, 2019 cka), Gdan sk, Poland; the Stein Eye Institute (Hoffer), University of California, Los Angeles, and St. Mary’s Eye Center (Hoffer), From Hygeia Clinic (Kanclerz, Przew1o Santa Monica, California, USA; the Department of Ophthalmology (Rozema), Antwerp University Hospital, Edegem, and Faculty of Medicine and Health Sciences (Rozema), University of Antwerp, Belgium; the IRCCS Fondazione Bietti (Savini), Rome, Italy. The contribution of IRCCS GB Bietti Foundation was supported by the Italian Ministry of Health and Fondazione Roma, Italy. sk, Poland. Email: [email protected]. Corresponding author: Piotr Kanclerz, MD, PhD, ul. Jaskowa Dolina 57, 80-286 Gdan Q 2019 ASCRS and ESCRS Published by Elsevier Inc.
0886-3350/$ - see frontmatter https://doi.org/10.1016/j.jcrs.2019.07.002
1620
OCT VS OLCR
Figure 1. Optical biometry results obtained with the SD-OCT device in a healthy patient with a clear lens (AL Z axial length; ACD Z anterior chamber depth; LT Z lens thickness; CCT Z central corneal thickness).
advantage to clinicians because they would have to purchase only one device rather than two. The purpose of this study was to examine the repeatability and reproducibility of the measurements provided by the new Revo NX optical biometer and compare them with those of a validated optical biometer,10 the Lenstar LS 900, which is based on optical low-coherence reflectometry (OLCR). PARTICIPANTS AND METHODS This prospective study enrolled volunteers with healthy eyes at Hygeia Clinic, Gda nsk, Poland, between July 2018 and August 2018. The study adhered to the tenets of the Declaration of Helsinki for the use of human participants in biomedical research and was approved by the local research ethics board. All participants signed informed consent forms after receiving an explanation of the purpose of the study. Exclusion criteria were any ocular disease (including cataract), corrected distance visual acuity worse than 20/25, previous ocular surgery, and trauma. Before being enrolled, all eyes had a complete ophthalmologic examination consisting of subjective refraction, noncontact tonometry, slitlamp evaluation, and ophthalmoscopy. Instruments The Revo NX SD-OCT device has an axial resolution of 5 mm, transverse resolution of 12 mm, and single scan depth of 2.4 mm
and obtains 110 000 scans per second. A superluminescent laser diode (830 nm) is used as the light source. The optical biometry program within the device measures the axial length (AL), anterior chamber depth (ACD; measured from the epithelium to lens), lens thickness (LT), and central corneal thickness (CCT). For each measurement, it performs 10 B-scans to calculate a precise average value (Figure 1). The Lenstar LS-900 optical low-coherence reflectometry (OLCR) biometer is a superluminescent diode laser (820 nm) that measures the AL, LT, CCT, and aqueous depth. The device also provides keratometry values, pupil size, and corneal diameter measurements. To allow comparison of the ACD values determined by the 2 devices, the ACD for the OLCR device was calculated as the sum of aqueous depth and CCT.11
Measurement Technique Measurements were performed with the 2 devices in random order. Only right eyes were analyzed, and all scans were taken between 15:00 and 19:00. Two skilled operators, in random order, performed 3 scans with the SD-OCT device to obtain 3 measurements. Repeatability was calculated for each observer, while reproducibility was assessed by a change in observer. To assess the agreement between the SD-OCT device and the OLCR device, the first observer took 3 measurements consecutively with both devices.
Table 1. Repeatability outcomes for 3 biometry measurements by each of 2 observers obtained using the spectral-domain optical coherence tomography device. Parameter AL (mm) Observer 1 Observer 2 ACD (mm) Observer 1 Observer 2 LT (mm) Observer 1 Observer 2 CCT (mm) Observer 1 Observer 2
Mean ± SD (mm)
Sr
r
CoV (%)
ICC
23.37 G 1.07 23.37 G 1.06
0.01 0.02
0.04 0.04
0.06 0.07
0.986 0.985
3.40 G 0.36 3.41 G 0.36
0.01 0.02
0.04 0.05
0.43 0.49
0.961 0.956
4.09 G 0.37 4.09 G 0.38
0.03 0.03
0.08 0.07
0.67 0.62
0.931 0.938
550.51 G 30.51 549.51 G 31.39
2.58 2.68
7.16 7.42
0.47 0.49
0.922 0.920
ACD Z anterior chamber depth; AL Z axial length; CCT Z central corneal thickness; CoV Z coefficient of variation; ICC Z intraclass correlation coefficients; LT Z lens thickness; r Z repeatability limit; Sr Z repeatability
Volume 45 Issue 11 November 2019
1621
OCT VS OLCR
Table 2. Reproducibility outcomes for biometry measurements obtained using the spectral-domain optical coherence tomography device. Parameter
Mean Difference ± SD
SR
R
CoV (%)
ICC
AL (mm) ACD (mm) LT (mm) CCT (mm)
0.00 G 0.02 0.01 G 0.03 0.00 G 0.04 1.00 G 5.43
0.01 0.01 0.01 1.76
0.02 0.04 0.04 4.88
0.03 0.38 0.37 0.32
0.993 0.965 0.962 0.945
ACD Z anterior chamber depth; AL Z axial length; CCT Z central corneal thickness; CoV Z coefficient of variation; ICC Z intraclass correlation coefficients; LT Z lens thickness; R Z reproducibility limit; Sr Z reproducibility
Statistical Analysis Statistical analysis was performed using Prism software (GraphPad Software, Inc.) and MedCalc software (MedCalc Software bvba). The results are presented as the mean G SD. The repeatability, repeatability limit, reproducibility, and reproducibility limit were calculated as recommended by McAlinden et al.12 The coefficient of variation (CoV) was determined as the ratio of the repeatability/reproducibility to the mean value (lower CoV stands for better reliability).13 The intraclass correlation coefficient (ICC) was defined as the ratio of variance between subjects to the sum of the pooled intraparticipant variance and the interparticipant variance. An ICC less than 0.75 indicates poor agreement; an ICC from 0.75 to 0.90 is considered as moderate, and an ICC of 0.90 or above is considered high.14 Comparison between the devices was performed using correlation coefficients, the Bland-Altman method, and paired t tests given that all parameters were normally distributed in the Kolmogorov-Smirnov test. A P value less than .05 was considered statistically significant.
RESULTS The study enrolled 65 participants. Two participants were excluded from the study; no measure for LT could be obtained in 1 eye and 1 participant withdrew consent for participation. Therefore, the results of 63 participants (44 women) were analyzed. The mean age was 41.8 G 13.8 years. The mean spherical equivalent refraction was 0.11 G 2.02 diopters (D) (range 5.38 to C4.63 D). The AL, ACD, LT, and CCT measurements obtained with SD-OCT device had high repeatability for both observers (Table 1). The CoV was less than 0.7% for all
parameters, while the ICC was higher than 0.92 for all parameters. Table 2 shows the reproducibility of the outcomes measured by the SD-OCT device. The reproducibility limit values were excellent, showing low variability. There was a statistically significant difference in AL, ACD, and LT measurements between the SD-OCT device and OLCR device (P ! .01). The mean AL and ACD measurements were higher with the SD-OCT device (Table 3). The CCT measurements were not significantly different between the devices (P Z .07). The correlation between the systems was very strong for AL (r Z 0.9997), ACD (r Z 0.9953), LT (r Z 0.9626), and CCT (r Z 0.9922). Bland-Altman analysis showed narrow 95% limits of agreement for AL, ACD, LT, and CCT measurements. Figures 2 to 5 show the Bland-Altman plots for these parameters. DISCUSSION Information about the repeatability and reproducibility of measurements and agreement with validated devices is essential when a new instrument becomes commercially available. Our data show that the measurements provided by the new Revo NX SD-OCT optical biometer offer high repeatability and reproducibility and a very strong correlation with those by the Lenstar LS 900 OLCR device. Although the agreement was good for all parameters except AL, a statistically significant difference was detected for all
Table 3. Mean values, mean differences, 95% LoA, and correlation for differences between the SD-OCT device and the OLCR device. Parameter/Device AL (mm) SD-OCT OLCR ACD (mm) SD-OCT OLCR LT (mm) SD-OCT OLCR CCT (mm) SD-OCT OLCR
Mean ± SD
Mean Difference ± SD
95% LoA
r Value
P Value
23.38 G 1.07 23.27 G 1.07
0.11 G 0.02
0.06, 0.15
0.9997
!.00001
3.39 G 0.36 3.34 G 0.36
0.05 G 0.04
0.02, 0.12
0.9953
!.00001
4.09 G 0.37 4.12 G 0.37
0.03 G 0.10
0.23, 0.17
0.9626
!.00001
550.51 G 30.51 551.34 G 30.50
0.83 G 3.57
7.83, 6.17
0.9922
!.00001
ACD Z anterior chamber depth; AL Z axial length; CCT Z central corneal thickness; LoA Z limits of agreement; LT Z lens thickness; OLCR Z optical lowcoherence reflectometry; r Z correlation coefficient; SD-OCT Z spectral-domain optical coherence tomography
Volume 45 Issue 11 November 2019
1622
OCT VS OLCR
Figure 2. Bland-Altman plot of agreement in axial length measurements between SD-OCT and OLCR devices (OLCR Z optical low-coherence reflectometry device; SD-OCT Z spectral-domain optical coherence tomography device).
Figure 3. Bland-Altman plot of agreement in anterior chamber depth measurements between SD-OCT and OLCR devices (OLCR Z optical low-coherence reflectometry device; SDOCT Z spectral-domain optical coherence tomography device).
of the 4 parameters evaluated in this study (AL, ACD, CCT LT). The largest dissimilarities were observed for the AL (mean difference C0.11 G 0.02 mm) and ACD (mean difference C0.05 G 0.04 mm). Such differences have not only statistical significance but also clinical significance and should not be underestimated; a shift of 0.1 mm in AL measurement transposes to approximately a 0.27 D change in the spectacle plane (for normal eye dimensions).15 Variations in biometric parameters have been reported with many optical biometers.16–25 In a study by Goebels et al.,21 AL values obtained with partial coherence interferometry (PCI) were on average 0.07 mm higher than those obtained with OLCR. Ortiz et al.18 reported that the AL determined with PCI was on average 0.04 mm shorter than that obtained with OLCR biometry. In a meta-analysis by Rozema et al.,10 the Lenstar OLCR device measurements were on average 0.02 G 0.01 mm (SD) higher than the IOLMaster PCI measurements, which was significantly equal. Table 4 shows the results in studies reporting differences in AL and ACD optical biometry measurements.1–3,7,16,18,21–25 The differences between the devices in our study could presumably be associated with the use of a dissimilar group refractive index to convert the optical path length into axial distances.26 Other potential reasons include the minimally different optical wavelength used by the devices or a dissimilar technique for automatic structure
border detection. These issues are likely a matter of calibration and should play no major role in the usefulness of the device. The difference in AL and ACD values in our study requires specific constant optimization for IOL power theoretical formulas that include them as predictors of the IOL position. This study and the SD-OCT optical biometer have limitations. First, we did not evaluate patients with eye disease, including cataract; therefore, our findings can be applied only to healthy eyes. Second, keratometric measurements and IOL calculation were not available in the device at the time of the study. Nevertheless, we believe this feature is an interesting development of OCT because most clinics likely have a validated keratometry device and theoretically could use the obtained data in IOL calculation formulas. Future software versions of this device enabling keratometric measurements will require further validation. In conclusion, the new Revo NX SD-OCT optical biometer provided repeatable and reproducible AL, ACD, LT, and CCT measurements. The results obtained with the SD-OCT biometer had a very strong correlation with those obtained with the Lenstar OLCR device; however, AL measurements, and to a lesser extent ACD measurements, cannot be considered interchangeable.
Figure 4. Bland-Altman plot of agreement in lens thickness measurements between SD-OCT and OLCR devices (OLCR Z optical low-coherence reflectometry device; SD-OCT Z spectral-domain optical coherence tomography device).
Figure 5. Bland-Altman plot of agreement in central corneal thickness measurements between SD-OCT and OLCR devices (OLCR Z optical low-coherence reflectometry device; SDOCT Z spectral-domain optical coherence tomography device).
Volume 45 Issue 11 November 2019
1623
OCT VS OLCR
Table 4. Studies reporting differences in AL and ACD measurements by optical biometry devices. Differences are presented as IOLMaster minus other device (mm). The IOLMaster 500 was used for all except 1 study,7 in which the 700 version was used. Axial Length (mm) Study*/Year Buckhurst 200923 Hoffer 20101 Rabsilber 201025 Epitropoulos 201416 Goebels 201521 Goebels 201521 Hoffer 20167 Hoffer 20162 Hoffer 20163 Whang 201822 Ortiz 201918 Ortiz 201918 Muzyka-Wozniak 201924
Anterior Chamber Depth (mm)
Device
n
Mean Diff
95% LoA
Mean Diff
Lenstar LS 900 Lenstar LS 900 Lenstar LS 900 Lenstar LS 900 Lenstar LS 900 OA-2000 Lenstar LS 900 AL-Scan Aladdin Argos Lenstar LS 900 Aladdin Pentacam AXL
112 50 100 105 138 138 183 86 60† 153 197 197 97
0.01 G 0.02 0.02 0.0 G 0.08z 0.03 G 0.03 0.02 G 0.03 0.05 G 0.09 0.01 0.0z 0.01z 0.03 0.04 G 0.26 0.04 G 0.13 0.01 G 0.04
d 0.03, 0.01 d 0.08, 0.03 d d 0.06, 0.03 0.03, 0.03 0.12, 0.09 d d d 0.01, 0.02
0.1 G 0.02 0.15 0.05 G 0.11 0.1 G 0.18 0.1 G 0.17 0.18 G 0.21 0.03 0.13 0.16 G 0.29 0.01z 0.13 G 0.23 0.1z 0.01 G 0.1
95% LoA d 0.21, 0.08 d 0.45, 0.25 d 0.09, 0.04 0.18, 0.44 0.78, 0.42 d d d 0.04, 0.02
Means G SD Diff Z difference; LoA Z limits of agreement *First Author † States cohort z Not statistically significant
WHAT WAS KNOWN Optical methods are considered to be the gold standard for preoperative biometry. At present, it is necessary to buy a separate optical coherence tomographer (OCT) device for anterior and posterior segment imaging and another for optical biometry.
WHAT THIS PAPER ADDS Optical biometry can be used in a new anterior and posterior segment spectral-domain (SD) OCT device. The new SD-OCT biometer provided repeatable and reproducible measurements of axial length, anterior chamber depth, lens thickness, and central corneal thickness. The results obtained with the new biometer showed a very strong correlation with a validated optical biometer. However, they cannot be considered interchangeable.
REFERENCES 1. Hoffer KJ, Shammas HJ, Savini G. Comparison of 2 laser instruments for measuring axial length. J Cataract Refract Surg 2010; 36:644–648; erratum, 1066 2. Hoffer KJ, Savini G. Comparison of AL-Scan and IOLMaster 500 partial coherence interferometry optical biometers. J Refract Surg 2016; 32:694–698 3. Hoffer KJ, Shammas HJ, Savini G, Huang J. Multicenter study of optical low-coherence interferometry and partial-coherence interferometry optical biometers with patients from the United States and China. J Cataract Refract Surg 2016; 42:62–67 4. Huang J, Savini G, Hoffer KJ, Chen H, Lu W, Hu Q, Bao F, Wang Q. Repeatability and interobserver reproducibility of a new optical biometer based on swept-source optical coherence tomography and comparison with IOLMaster. Br J Ophthalmol 2017; 101:493–498 5. Savini G, Negishi K, Hoffer KJ, Schiano Lomoriello D. Refractive outcomes of intraocular lens power calculation using different corneal power measurements with a new optical biometer. J Cataract Refract Surg 2018; 44:701–708
€ller M, Kohnen T. Com6. Shajari M, Cremonese C, Petermann K, Singh P, Mu parison of axial length, corneal and anterior chamber depth measurements of two recently introduced devices to a known biometer. Am J Ophthalmol 2017; 178:58–64 7. Hoffer KJ, Hoffmann PC, Savini G. Comparison of a new optical biometer using swept-source optical coherence tomography and a biometer using optical low-coherence reflectometry. J Cataract Refract Surg 2016; 42:1165–1172 8. Shammas HJ, Ortiz S, Shammas MC, Kim SH, Chong C. Biometry measurements using a new large-coherence–length swept-source optical coherence tomographer. J Cataract Refract Surg 2016; 42:50–61 9. Kanclerz P. Optical biometry in a commercially available anterior and posterior segment optical coherence tomography (OCT) device. Clin Exp Optom 2019 [Epub ahead of print] 10. Rozema JJ, Wouters K, Mathysen DGP, Tassignon M-J. Overview of the repeatability, reproducibility, and agreement of the biometry values provided by various ophthalmic devices. Am J Ophthalmol 2014; 158:1111–1120 11. Hoffer KJ. Definition of ACD [letter]. Ophthalmology 2011; 118:1484 12. McAlinden C, Khadka J, Pesudovs K. Precision (repeatability and reproducibility) studies and sample-size calculation [guest editorial]. J Cataract Refract Surg 2015; 41:2598–2604 13. Bland JM, Altman DG. Measurement error [statistics notes]. BMJ 1996; 312:1654 €ller R, Bu €ttner P. A critical discussion of intraclass correlation coeffi14. Mu cients. Stat Med 1994; 13:2465–2476 15. Olsen T. Calculation of intraocular lens power: a review. Acta Ophthalmol Scand 2007; 85:472–485 16. Epitropoulos A. Axial length measurement acquisition rates of two optical biometers in cataractous eyes. Clin Ophthalmol 2014; 8:1369–1376 17. Holzer MP, Mamusa M, Auffarth GU. Accuracy of a new partial coherence interferometry analyser for biometric measurements. Br J Ophthalmol 2009; 93:807–810 ~a V, Corrales MI, Ochoa M, Rodriguez CJ. 18. Ortiz A, Galvis V, Tello A, Vian Comparison of three optical biometers: IOLMaster 500, Lenstar LS 900 and Aladdin. Int Ophthalmol 2019; 39:1809–1818 19. Kongsap P. Comparison of a new optical biometer and a standard biometer in cataract patients. Eye Vis 2016; 3:27 20. Savini G, Hoffer KJ, Schiano-Lomoriello D. Agreement between lens thickness measurements by ultrasound immersion biometry and optical biometry. J Cataract Refract Surg 2018; 44:1463–1468 €ller M, Eppig T, Cayless A, Seitz B, Langenbucher A. 21. Goebels S, Pattmo Comparison of 3 biometry devices in cataract patients. J Cataract Refract Surg 2015; 41:2387–2393
Volume 45 Issue 11 November 2019
1624
OCT VS OLCR
22. Whang W-J, Yoo Y-S, Kang M-J, Joo C-K. Predictive accuracy of partial coherence interferometry and swept-source optical coherence tomography for intraocular lens power calculation. Sci Rep 2018; 8:13732 23. Buckhurst PJ, Wolffsohn JS, Shah S, Naroo SA, Davies LN, Berrow EJ. A new optical low coherence reflectometry device for ocular biometry in cataract patients. Br J Ophthalmol 2009; 93:949–953 24. Muzyka-Wozniak M, Oleszko A. Comparison of anterior segment parameters and axial length measurements performed on a Scheimpflug device with biometry function and a reference optical biometer. Int Ophthalmol 2019; 39:1115–1122 25. Rabsilber TM, Jepsen C, Auffarth GU, Holzer MP. Intraocular lens power calculation: clinical comparison of 2 optical biometry devices. J Cataract Refract Surg 2010; 36:230–234 26. Haigis W, Lege B, Miller N, Schneider B. Comparison of immersion ultrasound biometry and partial coherence interferometry for intraocular lens calculation according to Haigis. Graefes Arch Clin Exp Ophthalmol 2000; 238:765–773
Disclosures: Dr. Kanclerz receives non-financial support from Visim and Optopol Technology SA. To ensure accurate programming of his formulas, Dr. Hoffer licenses the registered trademark
Volume 45 Issue 11 November 2019
name HofferÒ to Carl Zeiss Meditec AG (IOLMasters), Haag-Streit AG (LenStar/EyeStar), Heidelberg Engineering, Inc. (Anterion), €te Movu, Inc. (Argos), Nidek, Inc. (AL-Scan), Oculus Optikgera GmbH (Pentacam AXL), Tomey Corp. (OA-2000), Topcon Europe Medical B.V./Visia Imaging S.r.l. (Aladdin), Ziemer Ophthalmic Systems AG (Galilei G6), and all A-scan biometer manufacturers. None of the other authors has a financial or proprietary interest in any material or method mentioned.
First author: Piotr Kanclerz, MD, PhD sk, Poland Hygeia Clinic , Gdan