Ophthalmic Surgery, Lasers and Imaging Retina

Clinical Science 

Reliability and Reproducibility of Pigment Epithelial Detachment Volume Measurements in AMD Using a New Tool: ReVAnalyzer

Avi Ohayon, MD; Oudy Semoun, MD; Violaine Caillaux, MD; Camille Jung, MD; Bruno Lay; Alexandra Miere, MD; Eric H. Souied, MD, PhD; Rocio Blanco-Garavito, MD

Abstract

BACKGROUND AND OBJECTIVE:

To describe and present the reliability and reproducibility of a new software, Retinal Volume Analyzer (ReVAnalyzer), for pigment epithelium detachment (PED) volume quantification.

PATIENTS AND METHODS:

This is a retrospective study including patients with PEDs secondary to exudative age-related macular degeneration (AMD). Macular volume scans on spectral-domain optical coherence tomography on enhanced depth imaging mode were performed in all eyes. Image batches were then exported in .xml format to the ReVAnalyzer software. A semiautomated PED volume measurement was performed by three independent readers (RBG, VC, OS) twice, at the beginning and end of a 15-day period. Bland-Altman assessment for agreement was used to compare intra- and interobserver observations.

RESULTS:

Twenty eyes of 20 patients presenting with PED were analyzed. Bland-Altman analysis indicated a good agreement between inter- and intraobserver measurements. The intraclass correlation coefficient for intraobserver PED volume measurements and between the three observers (interobserver) was greater than 0.99, demonstrating high reproducibility and consistency of the methodology.

CONCLUSIONS:

ReVAnalyzer is a reliable tool that can assist in the analysis of PED volume with high reproducibility. This type of specific retinal volume analysis can be of help for monitoring disease activity and therapeutic response in AMD.

[Ophthalmic Surg Lasers Imaging Retina. 2019;50:e242–e249.]

Abstract

BACKGROUND AND OBJECTIVE:

To describe and present the reliability and reproducibility of a new software, Retinal Volume Analyzer (ReVAnalyzer), for pigment epithelium detachment (PED) volume quantification.

PATIENTS AND METHODS:

This is a retrospective study including patients with PEDs secondary to exudative age-related macular degeneration (AMD). Macular volume scans on spectral-domain optical coherence tomography on enhanced depth imaging mode were performed in all eyes. Image batches were then exported in .xml format to the ReVAnalyzer software. A semiautomated PED volume measurement was performed by three independent readers (RBG, VC, OS) twice, at the beginning and end of a 15-day period. Bland-Altman assessment for agreement was used to compare intra- and interobserver observations.

RESULTS:

Twenty eyes of 20 patients presenting with PED were analyzed. Bland-Altman analysis indicated a good agreement between inter- and intraobserver measurements. The intraclass correlation coefficient for intraobserver PED volume measurements and between the three observers (interobserver) was greater than 0.99, demonstrating high reproducibility and consistency of the methodology.

CONCLUSIONS:

ReVAnalyzer is a reliable tool that can assist in the analysis of PED volume with high reproducibility. This type of specific retinal volume analysis can be of help for monitoring disease activity and therapeutic response in AMD.

[Ophthalmic Surg Lasers Imaging Retina. 2019;50:e242–e249.]

Introduction

Age-related macular degeneration (AMD) is the leading cause of blindness in patients older than 65 years of age in industrialized countries.1–3 Retinal thickness and volume analysis by optical coherence tomography (OCT) have become a reliable way of monitoring macular disease activity. Nevertheless, retinal volume can consist of several separate anatomical sections that can vary independently and cannot always be analyzed as a whole. There are different quantitative and qualitative ways to analyze the activity of a neovascular lesion with OCT, of which, the most common parameters analyzed are central macular thickness, total macular volume, presence of subretinal fluid (SRF) or intraretinal fluid (IRF), presence of subretinal hyperreflective material (SHRM), and pigment epithelial detachment (PED).4,5 Several retinal disorders are found to be associated with PEDs, such as central serous chorioretinopathy,6 inflammatory retinal diseases, and neoplastic etiologies, but the most common association is AMD in its different forms.7–12 PEDs develop in more than 80% of eyes with exudative AMD,7,13–15 and their size may vary in the same subject depending on disease activity. Calculating PED volume on follow-up for monitoring therapeutic response could possibly lead to the development of a separate biomarker of disease.16

The main limiting issue regarding volume calculation by spectral-domain OCT (SD-OCT) is the segmentation error done by the automatic segmentation software. Tilleul et al.17 described in a retrospective analysis of naïve exudative AMD patients that inner limiting membrane was correctly located by SD-OCT in 100% of images, whereas Bruch's membrane was correctly located by SD-OCT in 51.8% of images, with more frequent misallocations in cases of PED. Such errors in the automated detection of Bruch's membrane induce a wrong quantification of retinal thickness and might lead to erroneous decisions of re-treatment. To date, to the best of our knowledge, there are no readily available SD-OCT devices that can overcome the segmentation inaccuracy when dealing with PED without an intervention of an additional external software algorithm or without manual segmentation correction.

Penha et al.18 demonstrated high reproducibility of an algorithm software for the qualitative and quantitative assessment of PEDs in SD-OCT scans (Cirrus HD-OCT; Carl-Zeiss Meditec, Jena, Germany) that could be useful in following disease progression and response to treatment of vascularized PEDs. Nevertheless, this software was developed exclusively for this device.

We aimed to describe a new software, named ReVAnalyzer (Retinal Volume Analyzer; ADCIS, Saint-Contest, France), which may be used with other devices such as Spectralis HRA+OCT (Heidelberg Engineering, Heidelberg, Germany), and its reliability in PED volume measurements.

Patients and Methods

This retrospective study was performed in accordance with the Declaration of Helsinki and current French legislation and with the approval of our local ethics committee. We selected 20 subjects, older than age 50 years, presenting with either fibrovascular or mixed serous PEDs secondary to exudative AMD with subfoveal or juxtafoveal choroidal neovascularization (CNV), previously diagnosed on the basis of multimodal imaging. PED had to be of more than 250 μm in height (measured by SD-OCT) at its maximal height location, including extrafoveal locations for images to be selected and without a minimum of width. Eyes with PED width of more than 5.5 mm were excluded because the default scanning cube used in this study was 6 mm wide, aiming to analyze the whole PED volume. In order to get homogenous images, in this pilot study, without interference with the visibility of the retinal pigment epithelium (RPE), images with subfoveal fibrotic scar or hemorrhages were excluded.

All patients, taken from the ARI2 study,19 had undergone ophthalmological examination including Early Treatment Diabetic Retinopathy Study best-corrected visual acuity, slit-lamp examination, SD-OCT exam, fluorescein angiography, and indocyanine green angiography. SD-OCT on enhanced depth imaging mode (Spectralis HRA+OCT; Heidelberg Engineering, Heidelberg, Germany) was performed, with an acquisition of a 30° retinal volume scan consisting of 16 lines centered on the fovea with an average of nine frames per scan. This information was anonymized and exported in .xml format to the ReVAnalyzer software. This proprietary software was installed on a personal computer alongside the HEYEX image viewing platform (Heidelberg Engineering, Heidelberg, Germany) for image transportation. Image scale was fully preserved after exporting in .xml format. A semiautomatic analysis of PED volume was performed by three independent readers (RBG, VC, OS), twice, on a single scan, at the beginning and end of a 15-day period. A total of 16 slabs per scan were analyzed for each eye. The software (Figure 1) allows for semiautomated delineation of PED borders on a two-dimensional scan basis, after which automatic calculations were performed to present PED volume results in cubic millimeters (Figure 2). More specifically, after uploading the .xml file by the ReVAnalyzer software, the user interface (Figure 1) showed all 16 slabs for a particular scan, and then in the control panel, the “PED” green tab was chosen before beginning measurements. Alignment began manually by the reader along the RPE with the help of the mouse cursor, and the software detected automatically the Bruch's membrane (it detects a straight line with a homogenous reflectivity) to create a closed form. On each macular volume scan, running the software takes approximately 10 minutes (Video 1, below). The software used a triangulation process that connects artificially each slab to the next one and to the previous one for three-dimensional (3-D) reconstruction of the PED, volume was measured in three axes (X, Y, and Z), and the total volume was given in mm3 (Figure 2). We considered that the resolution in the three axes of the 3-D space were X = 5.5 μm/pixel, Y = 3.9 μm/pixel, and Z = 114.95 μm/pixel. Regarding statistical analysis, linear regression analysis was done to evaluate the correlation between the measurements (graph, coefficient regression calculation, and P values). The normality of the difference between measurements was assessed using the Shapiro-Wilk W test before performing reproducibility analysis. Bland-Altman assessment for an agreement was used to compare intra- and interobserver observations. Intraclass correlation coefficient (ICC) and 95% confidence interval were used to estimate the agreement between individual measurements of all readers.

Software user interface, one slab analysis. All 16 slabs for a particular scan are presented in the lower panel. Then, in the control panel, the “PED” green tab is chosen before beginning measurements. Alignment begins manually by the reader along the retinal pigment epithelium with the help of the mouse cursor, and the software detects automatically the Bruch's membrane to create a closed form. Before delineation (A). After delineation, pigment epithelium detachment content colored green (B).

Figure 1.

Software user interface, one slab analysis. All 16 slabs for a particular scan are presented in the lower panel. Then, in the control panel, the “PED” green tab is chosen before beginning measurements. Alignment begins manually by the reader along the retinal pigment epithelium with the help of the mouse cursor, and the software detects automatically the Bruch's membrane to create a closed form. Before delineation (A). After delineation, pigment epithelium detachment content colored green (B).

Depiction of three-dimensional pigment epithelium detachment (PED) volume representation. The PED is measured in three axes (X, Y, and Z), and the total volume is given in mm3.

Figure 2.

Depiction of three-dimensional pigment epithelium detachment (PED) volume representation. The PED is measured in three axes (X, Y, and Z), and the total volume is given in mm3.

Results

Twenty eyes of 20 patients presenting with PED secondary to exudative AMD were analyzed. The mean age was 78.04 years ± 7.34 years (range: 62.82 years to 92.50 years), 15 patients were women, and all were classified as occult CNV. In all cases, the PED volume was easily delineated. All PED volumes measured by the three viewers are shown in Table 1. These PED volume results are presented in cubic millimeters, measured independently by the three viewers. PED volume measurements of the same macular volume scan were done twice by each viewer at an interval of 15 days.

PED Volume Measurements in Cubic Millimeters by Three Viewers (RBG, VC, OS)

Table 1:

PED Volume Measurements in Cubic Millimeters by Three Viewers (RBG, VC, OS)

Linear regression calculations between measurements were excellent (Table 2, Figure 3). Bland-Altman plots were drawn to analyze intra- and interobserver reproducibility. Points of the plots were grouped and closed to the mean, indicating a good agreement between the measurements (Figure 4). One patient's measurements were outside limits of agreements. This could be explained by the presence of multiple PEDs, which interfered with manual and automatic alignments between readers.

Linear Regression Between Measurements by Three Viewers (RBG, VC, OS)

Table 2:

Linear Regression Between Measurements by Three Viewers (RBG, VC, OS)

Linear regression analysis plots of intra- and interobserver measurements of the three readers (RBG, VC, OS).

Figure 3.

Linear regression analysis plots of intra- and interobserver measurements of the three readers (RBG, VC, OS).

Bland-Altman analysis plots of intra- and interobserver measurements of the three readers (RBG, VC, OS).

Figure 4.

Bland-Altman analysis plots of intra- and interobserver measurements of the three readers (RBG, VC, OS).

The ICC for intraobserver PED volume measurements and between the three observers (interobserver) were greater than 0.99, demonstrating a correct reproducibility and consistency of the methodology, as can be shown in Table 3. The coefficient of repeatability calculations for measurements within each reader was low, suggesting high reproducibility between measurements (Table 4).

Intra- and Interobserver Intraclass Correlation Coefficient by Three Viewers (RBG, VC, OS)

Table 3:

Intra- and Interobserver Intraclass Correlation Coefficient by Three Viewers (RBG, VC, OS)

Coefficient of Repeatability Between Measurements by Three Viewers (RBG, VC, OS)

Table 4:

Coefficient of Repeatability Between Measurements by Three Viewers (RBG, VC, OS)

Discussion

In this study, we describe and present the reliability and reproducibility of a new software for PED volume measurement, called ReVAnalyzer. Measuring PED volume could be important for managing several retinal diseases, especially exudative AMD, in which PED height is for some authors a marker of disease activity.20–23

Schmidt-Erfurth et al.20 identified PED as the most relevant parameter reflecting progressive disease activity in exudative AMD, providing the link between the primary lesion activity morphology and the consecutive loss of retinal function. In their study, PED activation emerged as the primary underlying trigger for loss of vision in a pro re nata regimen. On the other hand, other studies showed a limited responsiveness of PED volume to treatment such as in the study of Keane et al.,24 which demonstrated a slow and incomplete resolution of PED despite effective and well-conducted antiangiogenic treatment and regimen. According to Jaffe et al.25 in the Comparison of Age-related Macular Degeneration Treatments Trials, although reductions in retinal fluid were impressive, PED volume tended to remain unchanged or to regress only slowly.

A few ongoing studies regarding PED changes measure the PED height at one point (foveal location) instead of its volume. We believe that PED height falsely represents the entire lesion size and fails to detect a global change in volume. As shown by Penha et al., both PED volume and area measurements may be useful in determining when to re-treat with antiangiogenic therapy.23 ReVAnalyzer software offers a much more accurate substitute to height measurements when dealing with volume quantification.

Lamin et al.26 demonstrated an automatic OCT segmentation software named Orion (Voxeleron; San Francisco, CA) to evaluate longitudinally volume changes in inner and outer retinal layers in early and intermediate AMD but did not include cases of PEDs. Other previously published algorithms were prone to artifacts.27,28 Ahlers et al.29 were the first to report a fully automated segmentation algorithm to obtain 3-D measurements in fibrovascular PEDs using the Cirrus SD-OCT imaging system; however, the RPE fit algorithm they used was not designed to identify the anatomic boundaries required for accurate volume and area measurements. Penha et al.18 showed a fully automated, highly reproducible algorithm for measuring PEDs. Although semiautomated segmentation might be time-consuming compared with automated software, it should be more accurate because segmentation is made by the observer.

In this paper, we describe and present the reliability and reproducibility of only the PED volume measurements. Theoretically, this semiautomated segmentation software is able to measure any volume in the chorioretinal tissue such as SRF, IRF (by automatic detection of black pixels after starting manual alignment), SHRM, fibrotic scars, or any lesion between segmentation lines (Video 2, below). Naturally, further study is necessary to assess the reproducibility of this software in these particular conditions. Our results showed a high intra- and interobserver reliability, hence this novel software presented here seems to be a promising semiautomated tool in PED volume analysis and could serve as a basis for re-treatment decisions in certain clinical cases.

There might be a limitation with measuring ill-defined PEDs or multiple PEDs as shown earlier with one patient with three PEDs whose measurements were outside of limits of agreements. Hence, the software may not be able to work properly in every case of PED and thus has a limited functionality in the clinic. Another technical limitation could be a misalignment of an extremely curved Bruch's membrane. However, these limitations could be corrected manually after automatic alignment. However, data analysis in our study should be considered carefully because the refractive index of the fluid inside the PED might interfere with measurements.30 Currently, this software is compatible only with the HEYEX software, but theoretically it is able to analyze any type of slab under the restriction that the image resolution acquisition is known in order to generate measurements in real-world units.

In conclusion, ReVAnalyzer is a reliable tool to evaluate and quantify PED volume with high precision. Retinal volume analysis can be helpful for monitoring disease activity and therapeutic response in AMD.

References

  1. SanGiovanni JP, Chew EY, Clemons TE, Age-Related Eye Disease Study Research Group et al. The relationship of dietary carotenoid and vitamin A, E, and C intake with age-related macular degeneration in a case-control study: AREDS Report No. 22. Arch Ophthalmol. 2007;125(9):1225–1232. doi:10.1001/archopht.125.9.1225 [CrossRef]17846363
  2. Bressler NM. Age-related macular degeneration is the leading cause of blindness. JAMA. 2004;291(15):1900–1901. doi:10.1001/jama.291.15.1900 [CrossRef]15108691
  3. Klein R, Klein BE, Jensen SC, Meuer SM. The five-year incidence and progression of age-related maculopathy: The Beaver Dam Eye Study. Ophthalmology. 1997;104(1):7–21. doi:10.1016/S0161-6420(97)30368-6 [CrossRef]9022098
  4. Nagiel A, Sarraf D, Sadda SR, et al. Type 3 neovascularization: Evolution, association with pigment epithelial detachment, and treatment response as revealed by spectral domain optical coherence tomography. Retina. 2015;35(4):638–647. doi:10.1097/IAE.0000000000000488 [CrossRef]25650713
  5. Coscas F, Coscas G, Souied E, Tick S, Soubrane G. Optical coherence tomography identification of occult choroidal neovascularization in age-related macular degeneration. Am J Ophthalmol. 2007;144(4):592–599. doi:10.1016/j.ajo.2007.06.014 [CrossRef]17698019
  6. Yannuzzi NA, Mrejen S, Capuano V, Bhavsar KV, Querques G, Freund KB. A central hyporeflective subretinal lucency correlates with a region of focal leakage on fluorescein angiography in eyes with central serous chorioretinopathy. Ophthalmic Surg Lasers Imaging Retina. 2015;46(8):832–836. doi:10.3928/23258160-20150909-07 [CrossRef]26431298
  7. Mrejen S, Sarraf D, Mukkamala SK, Freund KB. Multimodal imaging of pigment epithelial detachment: A guide to evaluation. Retina. 2013;33(9):1735–1762. doi:10.1097/IAE.0b013e3182993f66 [CrossRef]23873168
  8. Spaide RF. Enhanced depth imaging optical coherence tomography of retinal pigment epithelial detachment in age-related macular degeneration. Am J Ophthalmol. 2009;147(4):644–652. doi:10.1016/j.ajo.2008.10.005 [CrossRef]19152869
  9. Do DV, Gower EW, Cassard SD, et al. Detection of new-onset choroidal neovascularization using optical coherence tomography: The AMD DOC Study. Ophthalmology. 2012;119(4):771–778. doi:10.1016/j.ophtha.2011.10.019 [CrossRef]22297028
  10. Tan ACS, Freund KB, Balaratnasingam C, Simhaee D, Yannuzzi LA. Imaging of pigment epithelial detachments with optical coherence tomography angiography. Retina. 2018;38(9):1759–1769. doi:10.1097/IAE.0000000000002016 [CrossRef]29293207
  11. Mukkamala SK, Costa RA, Fung A, Sarraf D, Gallego-Pinazo R, Freund KB. Optical coherence tomographic imaging of sub-retinal pigment epithelium lipid. Arch Ophthalmol. 2012;130(12):1547–1553. doi:10.1001/archophthalmol.2012.2491 [CrossRef]22892986
  12. Balaratnasingam C, Messinger JD, Sloan KR, Yannuzzi LA, Freund KB, Curcio CA. Histologic and optical coherence tomographic correlates in drusenoid pigment epithelium detachment in age-related macular degeneration. Ophthalmology. 2017;124(5):644–656. doi:10.1016/j.ophtha.2016.12.034 [CrossRef]28153442
  13. Chan CK, Meyer CH, Gross JG, et al. Retinal pigment epithelial tears after intravitreal bevacizumab injection for neovascular age-related macular degeneration. Retina. 2007;27(5):541–551. doi:10.1097/IAE.0b013e3180cc2612 [CrossRef]17558314
  14. Pauleikhoff D, Löffert D, Spital G, et al. Pigment epithelial detachment in the elderly. Clinical differentiation, natural course and pathogenetic implications. Graefes Arch Clin Exp Ophthalmol. 2002;240(7):533–538. doi:10.1007/s00417-002-0505-8 [CrossRef]12136282
  15. Clemens CR, Krohne TU, Charbel Issa P, et al. High-resolution optical coherence tomography of subpigment epithelial structures in patients with pigment epithelium detachment secondary to age-related macular degeneration. Br J Ophthalmol. 2012;96(8):1088–1091. doi:10.1136/bjophthalmol-2011-301415 [CrossRef]22694959
  16. Wolf A, Rüping J, Neubauer AS, et al. Alterations of vascular pigment epithelium detachments associated with age-related macular degeneration during upload with intravitreal ranibizumab. Retina. 2013;33(9):1843–1849. doi:10.1097/IAE.0b013e318287d9be [CrossRef]23598795
  17. Tilleul J, Querques G, Canoui-Poitrine F, Leveziel N, Souied EH. Assessment of a spectral domain OCT segmentation software in a retrospective cohort study of exudative AMD patients. Ophthalmologica. 2013;229(2):80–85. doi:10.1159/000342980 [CrossRef]
  18. Penha FM, Rosenfeld PJ, Gregori G, et al. Quantitative imaging of retinal pigment epithelial detachments using spectral-domain optical coherence tomography. Am J Ophthalmol. 2012;153(3):515–523. doi:10.1016/j.ajo.2011.08.031 [CrossRef]
  19. Blanco-Garavito R, Jung C, Uzzan J, et al. Aflibercept after ranibizumab intravitreal injections in exudative age-related macular degeneration: The ARI2 study. Retina. 2018;38(12):2285–2292.
  20. Schmidt-Erfurth U, Waldstein SM, Deak GG, Kundi M, Simader C. Pigment epithelial detachment followed by retinal cystoid degeneration leads to vision loss in treatment of neovascular age-related macular degeneration. Ophthalmology. 2015;122(4):822–832. doi:10.1016/j.ophtha.2014.11.017 [CrossRef]25578255
  21. Balaskas K, Karampelas M, Horani M, Hotu O, Keane P, Aslam T. Quantitative analysis of pigment epithelial detachment response to different anti-vascular endothelial growth factor agents in wet age-related macular degeneration. Retina. 2017;37(7):1297–1304. doi:10.1097/IAE.0000000000001342 [CrossRef]
  22. Chan CK, Abraham P, Sarraf D. High-dose ranibizumab therapy for vascularized pigment epithelial detachment. Eye (Lond). 2012;26(6):882–885. doi:10.1038/eye.2012.90 [CrossRef]
  23. Penha FM, Gregori G, Garcia Filho CA, de A, Yehoshua Z, Feuer WJ, Rosenfeld PJ. Quantitative changes in retinal pigment epithelial detachments as a predictor for retreatment with anti-VEGF therapy. Retina. 2013;33(3):459–466. doi:10.1097/IAE.0b013e31827d2657 [CrossRef]23400085
  24. Keane PA, Liakopoulos S, Ongchin SC, et al. Quantitative subanalysis of optical coherence tomography after treatment with ranibizumab for neovascular age-related macular degeneration. Invest Ophthalmol Vis Sci. 2008;49(7):3115–3120. doi:10.1167/iovs.08-1689 [CrossRef]18408176
  25. Jaffe GJ, Martin DF, Toth CA, et al. Comparison of Age-related Macular Degeneration Treatments Trials Research Group. Macular morphology and visual acuity in the comparison of age-related macular degeneration treatments trials. Ophthalmology. 2013;120(9):1860–1870. doi:10.1016/j.ophtha.2013.01.073 [CrossRef]23642377
  26. Lamin A, Oakley JD, Dubis AM, Russakoff DB, Sivaprasad S. Changes in volume of various retinal layers over time in early and intermediate age-related macular degeneration. Eye (Lond). 2019;33(3):428–434. doi:10.1038/s41433-018-0234-9 [CrossRef]
  27. Szkulmowski M, Wojtkowski M, Sikorski B, et al. Analysis of posterior retinal layers in spectral optical coherence tomography images of the normal retina and retinal pathologies. J Biomed Opt. 2007;12(4):041207. doi:10.1117/1.2771569 [CrossRef]17867796
  28. Jain N, Farsiu S, Khanifar AA, et al. Quantitative comparison of drusen segmented on SD-OCT versus drusen delineated on color fundus photographs. Invest Ophthalmol Vis Sci. 2010;51(10):4875–4883. doi:10.1167/iovs.09-4962 [CrossRef]20393117
  29. Ahlers C, Simader C, Geitzenauer W, et al. Automatic segmentation in three-dimensional analysis of fibrovascular pigmentepithelial detachment using high-definition optical coherence tomography. Br J Ophthalmol. 2008;92(2):197–203. doi:10.1136/bjo.2007.120956 [CrossRef]
  30. Wojtkowski M, Leitgeb R, Kowalczyk A, Bajraszewski T, Fercher AF. In vivo human retinal imaging by Fourier domain optical coherence tomography. J Biomed Opt. 2002;7(3):457–463. doi:10.1117/1.1482379 [CrossRef]12175297

PED Volume Measurements in Cubic Millimeters by Three Viewers (RBG, VC, OS)

Patient Number First Measurement (RBG) Second Measurement (RBG) First Measurement (VC) Second Measurement (VC) First Measurement (OS) Second Measurement (OS)
1 0.913 0.933 0.937 0.940 0.792 0.822
2 0.513 0.557 0.552 0.527 0.560 0.548
3 0.567 0.509 0.542 0.570 0.496 0.527
4 3.008 2.988 3.038 2.861 2.922 2.944
5 3.264 3.238 3.256 3.286 2.918 2.945
6 0.310 0.335 0.349 0.349 0.315 0.315
7 0.570 0.584 0.584 0.585 0.571 0.576
8 1.404 1.410 1.438 1.456 1.456 1.472
9 0.550 0.612 0.618 0.540 0.884 0.814
10 1.126 1.152 1.110 1.141 1.148 1.131
11 5.044 5.087 5.187 5.145 4.897 5.052
12 1.309 1.334 1.389 1.383 1.303 1.241
13 9.189 9.129 9.169 9.249 9.102 9.257
14 1.451 1.689 1.818 1.703 1.745 2.024
15 0.215 0.199 0.226 0.212 0.189 0.210
16 0.848 0.839 0.815 0.830 0.765 0.982
17 0.878 0.842 0.852 0.856 0.781 0.854
18 0.639 0.674 0.668 0.676 0.658 0.737
19 1.137 1.161 1.166 1.161 1.046 1.073
20 0.515 0.544 0.531 0.537 0.519 0.576

Linear Regression Between Measurements by Three Viewers (RBG, VC, OS)

Reader Measurements Regression Coefficient (r) P Value
RBG First to second 0.999 < .0001
VC First to second 0.999 < .0001
OS First to second 0.999 < .0001
RBG-VC First to first 0.999 < .0001
RBG-OS First to first 0.997 < .0001
VC-OS First to first 0.998 < .0001

Intra- and Interobserver Intraclass Correlation Coefficient by Three Viewers (RBG, VC, OS)

Intraobserver Reproducibility Interobserver Reproducibility
Reader ICC 95% CI Readers ICC 95% CI
RBG 0.999 0.999–0.999 RBC and VC 0.999 0.998–0.999
VC 0.999 0.999–0.999 RBC and OS 0.996 0.994–0.999
OS 0.998 0.997–0.999 VC and OS 0.998 0.996–0.999

Coefficient of Repeatability Between Measurements by Three Viewers (RBG, VC, OS)

Reader Measurements Coefficient of Repeatability
RBG First to second 0.12
VC First to second 0.11
OS First to second 0.17

Authors

From the Department of Ophthalmology, University Paris Est Creteil, Centre Hospitalier Intercommunal de Creteil, Creteil, France (AO, OS, VC, CJ, AM, EHS, RBG); Centre d'Investigation Clinique, Centre Hospitalier Intercommunal de Creteil, Creteil, France (CJ); and ADCIS, Saint-Contest, France (BL).

Mr. Bruno is the chief executive officer of ADCIS, which developed the ReVAnalyzer software. The remaining authors report no relevant financial disclosures.

Address correspondence to Avi Ohayon, MD, Department of Ophthalmology, Centre Hospitalier Intercommunal de Créteil, Université de Paris Est Créteil, 40 Avenue de Verdun, 94000 Créteil, France; email: aviohayonmd@gmail.com.

Received: October 06, 2018
Accepted: March 25, 2019

10.3928/23258160-20190905-16

Sign up to receive

Journal E-contents