Quantitative Microvascular Analysis With Wide-Field Optical Coherence Tomography Angiography in Eyes With Diabetic Retinopathy | Diabetic Retinopathy | JAMA Network Open | JAMA Network
[Skip to Navigation]
Figure 1.  Representative Images of Eyes With Different Diabetic Retinopathy Severities
Representative Images of Eyes With Different Diabetic Retinopathy Severities
Figure 2.  Perfusion Density Change and Discriminative Power in Stratified Diabetic Retinopathy Severities
Perfusion Density Change and Discriminative Power in Stratified Diabetic Retinopathy Severities

Receiver operating characteristic (ROC) curves for discriminating diabetic retinopathy severities using total perfusion density and capillary perfusion density. AUC indicates area under the ROC curve.

Table 1.  Characteristics of Study Participants by Diabetes and DR Status
Characteristics of Study Participants by Diabetes and DR Status
Table 2.  Result of Optical Coherence Tomographic Angiography Analysis in Participants by Diabetes and DR Status
Result of Optical Coherence Tomographic Angiography Analysis in Participants by Diabetes and DR Status
Table 3.  Comparison of Vascular Density Parameters Between DR Severity Groupsa
Comparison of Vascular Density Parameters Between DR Severity Groupsa
1.
Cheung  N, Mitchell  P, Wong  TY.  Diabetic retinopathy.   Lancet. 2010;376(9735):124-136. doi:10.1016/S0140-6736(09)62124-3 PubMedGoogle ScholarCrossref
2.
Wong  TY, Cheung  CMG, Larsen  M, Sharma  S, Simó  R.  Diabetic retinopathy.   Nat Rev Dis Primers. 2016;2:16012. doi:10.1038/nrdp.2016.12 PubMedGoogle ScholarCrossref
3.
Yau  JWY, Rogers  SL, Kawasaki  R,  et al; Meta-Analysis for Eye Disease (META-EYE) Study Group.  Global prevalence and major risk factors of diabetic retinopathy.   Diabetes Care. 2012;35(3):556-564. doi:10.2337/dc11-1909 PubMedGoogle ScholarCrossref
4.
Ting  DSW, Cheung  GCM, Wong  TY.  Diabetic retinopathy: global prevalence, major risk factors, screening practices and public health challenges: a review.   Clin Exp Ophthalmol. 2016;44(4):260-277. doi:10.1111/ceo.12696 PubMedGoogle ScholarCrossref
5.
Jia  Y, Bailey  ST, Hwang  TS,  et al.  Quantitative optical coherence tomography angiography of vascular abnormalities in the living human eye.   Proc Natl Acad Sci U S A. 2015;112(18):E2395-E2402. doi:10.1073/pnas.1500185112 PubMedGoogle ScholarCrossref
6.
Chua  J, Tan  B, Ang  M,  et al.  Future clinical applicability of optical coherence tomography angiography.   Clin Exp Optom. 2019;102(3):260-269. doi:10.1111/cxo.12854 PubMedGoogle ScholarCrossref
7.
Spaide  RF, Fujimoto  JG, Waheed  NK, Sadda  SR, Staurenghi  G.  Optical coherence tomography angiography.   Prog Retin Eye Res. 2018;64:1-55. doi:10.1016/j.preteyeres.2017.11.003 PubMedGoogle ScholarCrossref
8.
Ang  M, Tan  ACS, Cheung  CMG,  et al.  Optical coherence tomography angiography: a review of current and future clinical applications.   Graefes Arch Clin Exp Ophthalmol. 2018;256(2):237-245. doi:10.1007/s00417-017-3896-2 PubMedGoogle ScholarCrossref
9.
Dupas  B, Minvielle  W, Bonnin  S,  et al.  Association between vessel density and visual acuity in patients with diabetic retinopathy and poorly controlled type 1 diabetes.   JAMA Ophthalmol. 2018;136(7):721-728. doi:10.1001/jamaophthalmol.2018.1319 PubMedGoogle ScholarCrossref
10.
Durbin  MK, An  L, Shemonski  ND,  et al.  Quantification of retinal microvascular density in optical coherence tomographic angiography images in diabetic retinopathy.   JAMA Ophthalmol. 2017;135(4):370-376. doi:10.1001/jamaophthalmol.2017.0080 PubMedGoogle ScholarCrossref
11.
Agemy  SA, Scripsema  NK, Shah  CM,  et al.  Retinal vascular perfusion density mapping using optical coherence tomography angiography in normals and diabetic retinopathy patients.   Retina. 2015;35(11):2353-2363. doi:10.1097/IAE.0000000000000862 PubMedGoogle ScholarCrossref
12.
Fawzi  AA, Fayed  AE, Linsenmeier  RA, Gao  J, Yu  F.  Improved macular capillary flow on optical coherence tomography angiography after panretinal photocoagulation for proliferative diabetic retinopathy.   Am J Ophthalmol. 2019;206:217-227. doi:10.1016/j.ajo.2019.04.032 PubMedGoogle ScholarCrossref
13.
Hwang  TS, Gao  SS, Liu  L,  et al.  Automated quantification of capillary nonperfusion using optical coherence tomography angiography in diabetic retinopathy.   JAMA Ophthalmol. 2016;134(4):367-373. doi:10.1001/jamaophthalmol.2015.5658 PubMedGoogle ScholarCrossref
14.
Samara  WA, Shahlaee  A, Adam  MK,  et al.  Quantification of diabetic macular ischemia using optical coherence tomography angiography and its relationship with visual acuity.   Ophthalmology. 2017;124(2):235-244. doi:10.1016/j.ophtha.2016.10.008 PubMedGoogle ScholarCrossref
15.
Ishibazawa  A, Nagaoka  T, Takahashi  A,  et al.  Optical coherence tomography angiography in diabetic retinopathy: a prospective pilot study.   Am J Ophthalmol. 2015;160(1):35-44.e1. doi:10.1016/j.ajo.2015.04.021 PubMedGoogle ScholarCrossref
16.
Couturier  A, Mané  V, Bonnin  S,  et al.  Capillary plexus anomalies in diabetic retinopathy on optical coherence tomography angiography.   Retina. 2015;35(11):2384-2391. doi:10.1097/IAE.0000000000000859 PubMedGoogle ScholarCrossref
17.
Silva  PS, Dela Cruz  AJ, Ledesma  MG,  et al.  Diabetic retinopathy severity and peripheral lesions are associated with nonperfusion on ultrawide field angiography.   Ophthalmology. 2015;122(12):2465-2472. doi:10.1016/j.ophtha.2015.07.034 PubMedGoogle ScholarCrossref
18.
Silva  PS, Cavallerano  JD, Haddad  NMN,  et al.  Peripheral lesions identified on ultrawide field imaging predict increased risk of diabetic retinopathy progression over 4 years.   Ophthalmology. 2015;122(5):949-956. doi:10.1016/j.ophtha.2015.01.008 PubMedGoogle ScholarCrossref
19.
McNabb  RP, Grewal  DS, Mehta  R,  et al.  Wide field of view swept-source optical coherence tomography for peripheral retinal disease.   Br J Ophthalmol. 2016;100(10):1377-1382. doi:10.1136/bjophthalmol-2015-307480 PubMedGoogle ScholarCrossref
20.
Hirano  T, Kakihara  S, Toriyama  Y, Nittala  MG, Murata  T, Sadda  S.  Wide-field en face swept-source optical coherence tomography angiography using extended field imaging in diabetic retinopathy.   Br J Ophthalmol. 2018;102(9):1199-1203. doi:10.1136/bjophthalmol-2017-311358 PubMedGoogle ScholarCrossref
21.
Zhang  Q, Rezaei  KA, Saraf  SS, Chu  Z, Wang  F, Wang  RK.  Ultra-wide optical coherence tomography angiography in diabetic retinopathy.   Quant Imaging Med Surg. 2018;8(8):743-753. doi:10.21037/qims.2018.09.02 PubMedGoogle ScholarCrossref
22.
Pellegrini  M, Cozzi  M, Staurenghi  G, Corvi  F.  Comparison of wide field optical coherence tomography angiography with extended field imaging and fluorescein angiography in retinal vascular disorders.   PLoS One. 2019;14(4):e0214892. doi:10.1371/journal.pone.0214892 PubMedGoogle Scholar
23.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.   JAMA. 2013;310(20):2191-2194. doi:10.1001/jama.2013.281053PubMedGoogle ScholarCrossref
24.
Early Treatment Diabetic Retinopathy Study Research Group.  Fluorescein angiographic risk factors for progression of diabetic retinopathy. ETDRS report number 13.   Ophthalmology. 1991;98(5)(suppl):834-840. doi:10.1016/S0161-6420(13)38015-4 PubMedGoogle ScholarCrossref
25.
Hong  JM, Tan  B, Gupta  P,  et al.  Intrasession repeatability of quantitative metrics using widefield optical coherence tomography angiography (OCTA) in old community-dwelling Asian adults.   Invest Ophthalmol Vis Sci. 2019;60(9):3067.Google Scholar
26.
Wang  RK, An  L, Francis  P, Wilson  DJ.  Depth-resolved imaging of capillary networks in retina and choroid using ultrahigh sensitive optical microangiography.   Opt Lett. 2010;35(9):1467-1469. doi:10.1364/OL.35.001467 PubMedGoogle ScholarCrossref
27.
De Pretto  LR, Moult  EM, Alibhai  AY,  et al.  Controlling for artifacts in widefield optical coherence tomography angiography measurements of non-perfusion area.   Sci Rep. 2019;9(1):9096. doi:10.1038/s41598-019-43958-1 PubMedGoogle ScholarCrossref
28.
Koo  TK, Li  MY.  A guideline of selecting and reporting intraclass correlation coefficients for reliability research.   J Chiropr Med. 2016;15(2):155-163. doi:10.1016/j.jcm.2016.02.012 PubMedGoogle ScholarCrossref
29.
DeLong  ER, DeLong  DM, Clarke-Pearson  DL.  Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach.   Biometrics. 1988;44(3):837-845. doi:10.2307/2531595 PubMedGoogle ScholarCrossref
30.
Kolb  JP, Klein  T, Kufner  CL, Wieser  W, Neubauer  AS, Huber  R.  Ultra-widefield retinal MHz-OCT imaging with up to 100 degrees viewing angle.   Biomed Opt Express. 2015;6(5):1534-1552. doi:10.1364/BOE.6.001534 PubMedGoogle ScholarCrossref
31.
Hirano  T, Kitahara  J, Toriyama  Y, Kasamatsu  H, Murata  T, Sadda  S.  Quantifying vascular density and morphology using different swept-source optical coherence tomography angiographic scan patterns in diabetic retinopathy.   Br J Ophthalmol. 2019;103(2):216-221. doi:10.1136/bjophthalmol-2018-311942 PubMedGoogle ScholarCrossref
32.
Corvi  F, Pellegrini  M, Erba  S, Cozzi  M, Staurenghi  G, Giani  A.  Reproducibility of vessel density, fractal dimension, and foveal avascular zone using 7 different optical coherence tomography angiography devices.   Am J Ophthalmol. 2018;186:25-31. doi:10.1016/j.ajo.2017.11.011 PubMedGoogle ScholarCrossref
33.
Tan  B, Chua  J, Barathi  VA,  et al.  Quantitative analysis of choriocapillaris in non-human primates using swept-source optical coherence tomography angiography (SS-OCTA).   Biomed Opt Express. 2019;10(1):356-371. doi:10.1364/BOE.10.000356 PubMedGoogle Scholar
34.
Sun  C, Wang  JJ, Mackey  DA, Wong  TY.  Retinal vascular caliber: systemic, environmental, and genetic associations.   Surv Ophthalmol. 2009;54(1):74-95. doi:10.1016/j.survophthal.2008.10.003 PubMedGoogle ScholarCrossref
35.
Ikram  MK, Ong  YT, Cheung  CY, Wong  TY.  Retinal vascular caliber measurements: clinical significance, current knowledge and future perspectives.   Ophthalmologica. 2013;229(3):125-136. doi:10.1159/000342158 PubMedGoogle ScholarCrossref
36.
Pemp  B, Polska  E, Garhofer  G, Bayerle-Eder  M, Kautzky-Willer  A, Schmetterer  L.  Retinal blood flow in type 1 diabetic patients with no or mild diabetic retinopathy during euglycemic clamp.   Diabetes Care. 2010;33(9):2038-2042. doi:10.2337/dc10-0502 PubMedGoogle ScholarCrossref
37.
Fondi  K, Wozniak  PA, Howorka  K,  et al.  Retinal oxygen extraction in individuals with type 1 diabetes with no or mild diabetic retinopathy.   Diabetologia. 2017;60(8):1534-1540. doi:10.1007/s00125-017-4309-0 PubMedGoogle ScholarCrossref
38.
Suwan  Y, Fard  MA, Geyman  LS,  et al.  Association of myopia with peripapillary perfused capillary density in patients with glaucoma: an optical coherence tomography angiography study.   JAMA Ophthalmol. 2018;136(5):507-513. doi:10.1001/jamaophthalmol.2018.0776 PubMedGoogle ScholarCrossref
39.
Rosen  RB, Andrade Romo  JS, Krawitz  BD,  et al.  Earliest evidence of preclinical diabetic retinopathy revealed using optical coherence tomography angiography perfused capillary density.   Am J Ophthalmol. 2019;203:103-115. doi:10.1016/j.ajo.2019.01.012 PubMedGoogle ScholarCrossref
40.
Geyman  LS, Garg  RA, Suwan  Y,  et al.  Peripapillary perfused capillary density in primary open-angle glaucoma across disease stage: an optical coherence tomography angiography study.   Br J Ophthalmol. 2017;101(9):1261-1268. doi:10.1136/bjophthalmol-2016-309642 PubMedGoogle ScholarCrossref
41.
Liu  L, Jia  Y, Takusagawa  HL,  et al.  Optical coherence tomography angiography of the peripapillary retina in glaucoma.   JAMA Ophthalmol. 2015;133(9):1045-1052. doi:10.1001/jamaophthalmol.2015.2225 PubMedGoogle ScholarCrossref
42.
Rao  HL, Pradhan  ZS, Weinreb  RN,  et al.  Regional comparisons of optical coherence tomography angiography vessel density in primary open-angle glaucoma.   Am J Ophthalmol. 2016;171:75-83. doi:10.1016/j.ajo.2016.08.030 PubMedGoogle ScholarCrossref
43.
Sun  Z, Tang  FY, Wong  R,  et al.  OCT angiography metrics predict progression of diabetic retinopathy and development of diabetic macular edema: a prospective study.   Ophthalmology. 2019;126(12):1675-1684. doi:10.1016/j.ophtha.2019.06.016 PubMedGoogle Scholar
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    Ophthalmology
    January 17, 2020

    Análisis microvascular cuantitativo con angiografía por tomografía de coherencia óptica de campo amplio en ojos con retinopatía diabética

    Author Affiliations
    • 1Singapore Eye Research Institute, Singapore National Eye Centre, Singapore
    • 2Singapore Eye Research Institute–Nanyang Technological University Advanced Ocular Engineering (STANCE) Program, Singapore
    • 3Academic Clinical Program, Duke–National University of Singapore Medical School, Singapore
    • 4Department of Ophthalmology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore
    • 5Department of Ophthalmology, Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore
    • 6Department of Clinical Pharmacology, Medical University of Vienna, Vienna, Austria
    • 7Center for Medical Physics and Biomedical Engineering, Medical University of Vienna, Vienna, Austria
    • 8National University Health System, Singapore
    JAMA Netw Open. 2020;3(1):e1919469. doi:10.1001/jamanetworkopen.2019.19469
    Puntos claveEnglish 中文 (chinese)

    Pregunta  ¿Cuál es el rendimiento diagnóstico del análisis microvascular cuantitativo mediante angiografía por tomografía de coherencia óptica (optical coherence tomographic angiography, OCTA) de campo amplio en ojos con retinopatía diabética?

    Conclusiones  En este estudio de casos y controles de 49 ojos en 27 participantes de control y 76 ojos en 47 participantes con diabetes, las métricas microvasculares obtenidas de OCTA de campo amplio clasificaron con precisión la gravedad de la retinopatía diabética no proliferativa con alta sensibilidad.

    Significado  Los hallazgos sugieren que la OCTA de campo amplio puede ser útil para evaluar el estado microvascular en ojos con retinopatía diabética.

    Abstract

    Importance  Wide-field optical coherence tomographic angiography (OCTA) may provide insights to peripheral capillary dropout in eyes with diabetic retinopathy (DR).

    Objective  To describe the diagnostic performance of wide-field OCTA with and without large vessel removal for assessment of DR in persons with diabetes.

    Design, Setting, and Participants  This case-control study was performed from April 26, 2018, to April 8, 2019, at a single tertiary eye center in Singapore. Case patients were those with type 2 diabetes for more than 5 years and bilateral DR diagnosed by fundus imaging; control participants included those with no self-reported history of diabetes, a fasting glucose level within the normal range in the past year, and no ocular pathologic findings. A wide-field (12 × 12-mm2) fovea-centered scan was performed using a prototype swept source OCTA system. Retinal microvasculature was examined by separating the angiograms into large vessels, capillaries, and capillary dropout regions.

    Main Outcomes and Measures  Area under the receiver operating characteristic curve (AUC) for DR severity discrimination using wide-field vascular metrics. Retinal perfusion density (RPD), capillary perfusion density (CPD), large vessel density (LVD), and capillary dropout density (CDD) were calculated. Low-contrast regions were excluded from the calculation.

    Results  A total of 49 eyes in 27 control participants (17 male [63.0%]; mean [SD] age, 59.96 [7.63] years; age range, 44-79 years) and 76 eyes in 47 patients with diabetes (29 male [61.7%]; mean [SD] age, 64.36 [8.08] years; range, 41-79 years) were included. Among eyes in patients with diabetes, 23 were in those with diabetes but no DR, 25 in those with mild nonproliferative DR, and 28 in those with moderate to severe nonproliferative DR. There was no difference in RPD, CPD, LVD, and CDD between the control group and the group with diabetes and no DR. There was a stepwise decrease in RPD, CPD, and CDD in the diabetes with no DR, mild nonproliferative DR, and moderate to severe nonproliferative DR groups, whereas LVD was not associated with DR staging. The nonproliferative DR group had decreased RPD, CPD, and CDD compared with the control group. The CPD had higher AUCs than RPD for discriminating diabetes with nonproliferative DR (combined mild and moderate to severe nonproliferative DR) vs no DM (AUC, 0.92 [95% CI, 0.87-0.98] vs 0.89 [95% CI, 0.83-0.95], P = .01), diabetes with no DR vs mild nonproliferative DR (AUC, 0.81 [95% CI, 0.68-0.94] vs 0.77 [95% CI, 0.64-0.91], P = .18), and mild nonproliferative DR vs moderate to severe nonproliferative DR (AUC, 0.82 [95% CI, 0.71-0.94] vs 0.78 [95% CI, 0.65-0.91], P = .01) but similar AUCs for no DM vs diabetes with no DR. The total perfusion density and CPD in wide-field OCTA had better discriminative power than the central 6 × 6-mm2 field (CPD, 0.89 [95% CI, 0.83-0.95] vs 0.84 [95% CI, 0.77-0.92], P = .06; total perfusion density, 0.93 [95% CI, 0.87-0.98] vs 0.90 [95% CI, 0.83-0.96], P = .06).

    Conclusions and Relevance  The findings suggest that wide-field OCTA provides information on microvascular perfusion and may be useful for detecting predominant peripheral capillary dropout in eyes with nonproliferative DR. A vascular selectivity approach excluding the large vessels may improve the discriminative power for different stages of DR.

    Introduction

    Diabetic retinopathy (DR) is a common microvascular ocular complication of diabetes and is a leading cause of blindness in the working-age population.1,2 The population with diabetes globally is estimated to reach 366 million in 2030, with 34.6% having DR and 7% having vision-threatening DR.3,4

    Optical coherence tomography angiography (OCTA) has advanced the understanding of DR and is used as a tool for detecting capillary nonperfusion.5 Traditionally, nonperfusion areas could only be detected by fluorescein angiography, but OCTA is a noninvasive, rapid, and simple approach to provide a 3-dimensional representation of the retinal vascular network.6-8 Vascular metrics, such as properties of fovea avascular zone, perfusion density, vessel density, and vascular fractal dimension have been reported in the context of diabetes or DR.9-16

    Most current OCTA machines can image only a small field of view around the macula and provide little insight to the peripheral regions. Because vascular alterations occur predominantly in the peripheral area of the eye in DR,17-19 solely looking into a small field of view may result in misclassification (ie, eyes with poor peripheral perfusion may not be detected). Large field of view protocols have been recently adopted in several machines, and by stitching or adding an ocular lens, one can achieve a larger field of view up to a 20 × 20-mm2 field20-22; however, these methods have yet to be validated in larger populations.

    In this study, we investigated retinal microvascular changes using a wide-field OCTA scan (12 × 12-mm2 field) in patients with diabetes with or without DR compared with individuals without diabetes or ocular diseases. The wide-field images were further divided into a central 6 × 6-mm2 field and the remaining square annulus regions (Figure 1) to compare the wide-field protocol with traditional small field of view protocols. Localized capillary metrics were studied by uniformly dividing the entire field into 5 × 5 nonoverlapping blocks. We further hypothesized that large vessel removal from OCTA images would be associated with improved diagnostic performance of OCTA.

    Methods
    Study Participants

    This cross-sectional case-control study was approved by the SingHealth Centralized institutional review board, Singapore, and conducted in accordance with the Declaration of Helsinki.23 Written informed consent was obtained from all participants. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. Patients with diabetes and DR were compared with patients with diabetes without DR and patients without diabetes (controls). The study was performed from April 26, 2018, to April 8, 2019, at a single tertiary eye center, the Singapore National Eye Center (SNEC), Singapore. Participants meeting inclusion criteria were patients who had type 2 diabetes for more than 5 years and bilateral nonproliferative DR (NPDR) that was diagnosed by fundus imaging according to the Early Treatment Diabetic Retinopathy Study DR grading scale by the SNEC Ocular Reading Centre.24 Diabetes criteria were a known physician diagnosis of type 2 diabetes and receipt of oral hypoglycemic agents or insulin therapy. Exclusion criteria were glaucoma, age-related macular degeneration, ocular opacity, or history of panretinal photocoagulation. Inclusion criteria for the control participants were no self-reported history of diabetes, a fasting glucose level within the normal range in the previous year, and no ocular pathologic findings (including glaucoma, age-related macular degeneration, or ocular opacity).

    Another small group of participants (n = 12) was included from a study under the Prevention of and Intervention for Eye Diseases in the Elderly (PROVIDE) program.25 The results obtained for those participants were used to test the reproducibility of the measurements.

    Optical Coherence Tomography Angiography

    Optical coherence tomography angiography is a noninvasive imaging modality that provides 3-dimensional structural and angiographic information of the posterior pole of the retina. The prototype swept source OCT (SS-OCT) system (PlexElite 9000; Zeiss Meditec) uses a wavelength scanning laser (central wavelength [λc] = 1050 nm) as light source, and the spectral information is acquired by a photodetector. The system operation speed is dependent on the scanning rate of the swept source (100 000 amplitude scan [A-scan] per second), and the axial and lateral resolutions in tissue are 6.3 μm and 20 μm, respectively.

    All the participants were scanned by the same trained ophthalmic technician (E.L.). A wide-field area (12 × 12-mm2 field) centered at the fovea was scanned, and each data volume consisted of 500 A-scans and 500 B-scans. Each B-scan was repeated twice to generate OCTA images using an optical microangiography algorithm.26 Motion-related artifacts were minimized by an integrated line-scanning ophthalmoscope eye tracker during data acquisition. A review software, PlexElite Review Software, version 1.6 (Zeiss Meditec), provided automated segmentation of retinal layers, and the inner retina was defined from inner limiting membrane to 70 μm above the retinal pigment epithelium. En face maximum projection angiographic images including superficial and deep vascular plexuses were extracted for further analysis. In the present study, superficial and deep vascular plexuses were not separated because of insufficient performance of the projection artifact removal algorithm in the wide-field images. For the reproducibility analysis, 1 eye was randomly chosen from the 12 PROVIDE participants and scanned twice by the same ophthalmic technician with an interval of 5 minutes between scans.

    Image Analysis

    A custom MatLab algorithm (The MathWorks Inc) was developed for quantitative analysis. First, a combined Gabor and Hessian–based vessel filter was applied to enhance the contrast of large blood vessels, followed by a thresholding method to generate a binarization mask of the large vessels. Second, a sliding window scheme (window size, 3 × 3 mm2; sliding step, 1.5 mm) was used for segmenting capillaries. In each window, the threshold was empirically set to be 0.35 of the mean intensity of the large vessels. Third, all the segmented large vessel and capillary regions were set to 1 and the background was set to 0; the image contrast was then flipped by an inverse transform, and nonperfusion area was considered as 1. Intercapillary regions more than 0.36 mm2 were considered as capillary dropout regions. The foveal avascular zone and optic disc area were excluded manually from the capillary dropout calculation. The low-signal-related artifacts were filtered out based on the corresponding structural images.27 The following 4 vascular metrics were computed: (1) retinal total perfusion density (TPD; total perfused area per total imaged area), capillary perfusion density (CPD; capillary perfused area per total imaged area), large vessel density (LVD; large vessel area per total imaged area), and capillary dropout density (CDD; capillary dropout area per total imaged area). For TPD, CPD, and LVD, lower values suggest worse vascular disruption. For CDD, higher values suggest worse vascular disruption.

    To study localized vascular changes, the wide-field 12 × 12-mm2 images were separated into (1) a central 6 × 6-mm2 field and (2) the remaining square annulus consisting of 5 × 5 nonoverlapping uniform blocks with a size of 2.4 × 2.4 mm2. The 4 vascular metrics (TPD, CPD, LVD, and CDD) were calculated in each subregion.

    Statistical Analysis

    Data were analyzed using Stata, version 15.0 (StataCorp LLC) and R, version 3.5.0 (R Project for Statistical Computing). The repeatability of the measurement of each vascular metric was assessed using the intraclass correlation coefficient (ICC). The ICC values were defined as follows: less than 0.50 (poor repeatability), between 0.50 and 0.75 (moderate repeatability), between 0.75 and 0.90 (good repeatability), and greater than 0.90 (excellent repeatability).28

    We compared 4 vascular metrics between the 3 DR severity groups (no DR, mild NPDR, and moderate to severe NPDR). The sampling distribution of the mean pairwise difference was obtained by nonparametric bootstrapping (1000 replicates) of individuals as the resampling clusters. This approach was taken to account for skewness of the distribution of certain vascular parameters and between-eye correlation of measurements made in an individual. The false discovery rate was controlled at a level of 5% using the Benjamini-Hochberg procedure to adjust P values (2-sided P < .05 was considered statistically significant) for multiple pairwise comparisons across vascular parameters and combinations of DR severity groups. We also conducted a post hoc test of the linear orthogonal contrast, the result of which is referred to as P value for trend to access the monotonical increase or decrease of the vascular metrics in different retinopathy status. We assessed the performance of TPD and CPD in discriminating control eyes and eyes with NPDR using the receiver-operating characteristic (ROC) curve. Comparisons of areas under the ROC curve (AUCs) were performed using an algorithm proposed by DeLong et al.29

    Results
    Patient Characteristics

    A summary of patient characteristics is shown in Table 1. The initial quality check of wide-field OCTA excluded 30% to 40% of these scans because of multiple artifacts, such as low contrast, extensive motion artifacts, layer segmentation errors, and fixation problems. In the control group, a total of 49 eyes from 27 participants were included (17 male [63.0%]; mean [SD] age, 59.96 [7.63] years; age range, 44-79 years). In the group with diabetes, a total of 76 eyes from 47 participants were included (29 male [61.7%]; mean [SD] age, 64.36 [8.08] years; age range, 41-79 years). The DR severity was classified into 3 classes: no DR (23 eyes [30.2%]), mild NPDR (25 eyes [32.9%]), and moderate to severe NPDR (28 eyes [36.8%]). Higher glycated hemoglobin level was associated with DR severity. There was no difference in age, sex, serum glucose level, serum creatinine level, duration of diabetes, cholesterol level, high-density lipoprotein cholesterol level, triglyceride levels, low-density lipoprotein cholesterol level, and cholesterol ratio among groups, but systolic blood pressure was higher in the diabetes group.

    Repeatability

    The repeatability test was done for 12 patients with a mean (SD) age of 70.12 (6.68) years (range, 60-84 years). Six of the participants had diabetes, and 5 had systemic hypertension. The ICC scores were 0.80 for TPD, 0.81 for CPD, 0.80 for LVD, and 0.92 for CCD. In the cropped, central 6 × 6-mm2 field, ICC scores were lower (0.41 for TPD, 0.71 for CPD, and 0.49 for LVD). In the square annulus, area reproducibility was better, and the ICC scores were 0.84 for TPD, 0.81 for CPD, 0.90 for LVD, and 0.93 for CCD.

    Examples of fundus images, OCTA images, and processed images of patients in the control, no DR, mild NPDR, and moderate to severe NPDR groups are shown in Figure 1. The OCTA images showed decreased capillary perfusion and increased capillary dropout area associated with worsening DR severity, and the capillary dropout regions were located mainly in the peripheral parts of the angiogram.

    Quantitative Analysis

    The quantitative analysis is summarized in Table 2, and the P values from statistical tests are summarized in Table 3. In a comparison between any DR and no diabetes using wide-field scans, mean (SD) TPD and CPD decreased (TPD: 68.48% [11.15%] vs 85.91% [3.84%], P < .001; CPD: 51.51% [11.37%] vs 70.67% [3.67%], P < .001), whereas mean (SD) LVD and CDD increased (LVD: 16.94% [1.75%] vs 15.20% [1.00%], P < .001; CDD: 11.21% [8.39%] vs 0.42% [0.51%], P < .001). There was no difference in the OCTA metrics between the no DR and no diabetes groups. In a comparison of the mild NPDR group with the no DR groups, TPD and CPD decreased (TPD: –0.85 [95% CI, –1.33 to –0.30], P = .003; CPD: –0.89 [95% CI, –1.38 to –0.32], P = .002), whereas CDD increased (0.51 [95% CI, 0.07-0.91], P = .03). For the moderate to severe NPDR group vs the mild NPDR group, TPD and CPD decreased (TPD: –0.96 [95% CI, –1.49 to –0.45], P = .001; CPD: –1.01, [95 % CI, –1.46 to –0.59], P < .001), whereas CDD increased (1.07 [95% CI, 0.46-1.77], P = .002); there was no difference in LVD (0.67 [95% CI, 0.02-1.29], P = .06).

    The square annulus and the central 6 × 6-mm2 field showed similar differences between the any DR and no diabetes groups, but the square annulus showed greater OCTA differences than the central 6 × 6-mm2 field. Specifically, when comparing mild NPDR and no DR groups, CDD was not different in the central 6 × 6-mm2 field but was significantly increased in the square annulus (mean difference in units of SD, 0.51 [95% CI, 0.06-0.91]; P = .03), indicating that capillary dropout occurred predominantly in the peripheral regions during the early stages of NPDR.

    Qualitative Localized Dropout

    Figure 2A and B demonstrate localized capillary dropout associated with DR severity. The color in each of the 5 × 5-mm2 blocks represents the relative perfusion reduction. The capillary difference between the no DR and no diabetes groups was minimal. Mild NPDR was associated with subtle capillary dropout, and the change was uniformly distributed over the entire 12 × 12-mm2 field. The capillary dropout in the moderate to severe NPDR group mainly occurred in the periphery and most often in the temporal quadrant.

    Figure 2C shows the ROC curves for TPD and CPD for separation between DR severities and between controls and persons with NPDR. In a comparison between controls and those without DR, there was no difference between TPD and CPD in AUCs. By contrast, the AUCs of CPD were larger than those of TPD when comparing diabetes with no DR vs mild NPDR (AUC, 0.81 [95% CI, 0.68-0.94] vs 0.77 [95% CI, 0.64-0.91], P = .18), or mild NPDR vs moderate to severe NPDR (AUC, 0.82 [95% CI, 0.71-0.94] vs 0.78 [95% CI, 0.65-0.91], P = .01). Furthermore, CPD had a higher AUC than TPD when comparing eyes with NPDR (combined mild and moderate to severe NPDR) and control eyes (AUC, 0.92 [95% CI, 0.87-0.98] vs 0.89 [95% CI, 0.83-0.95], P = .01).

    The AUCs of TPD and CPP in the regions of 6 × 6-mm2 field and the remaining square annulus were also calculated. The TPD and CPP in wide-field OCTA had better discriminative power than the central 6 × 6-mm2 field (CPP: 0.89 [95% CI, 0.83-0.95] vs 0.84 [95% CI, 0.77-0.92], P = .06; TPD: 0.93 [95% CI, 0.87-0.98] vs 0.90 [95% CI, 0.83-0.96], P = .06), although this difference was not statistically significant. The AUCs were similar between the wide field (12 × 12-mm2 field) and the square annulus.

    Discussion

    In this article, we describe results of several quantitative microvascular metrics using a protocol of wide-field OCTA in patients with DR. The key findings of the present study are that wide-field OCTA may be useful to detect predominant peripheral capillary dropout in eyes with NPDR.

    Furthermore, we showed that separating the angiograms into large vessels, capillaries, and capillary dropout regions may improve the diagnostic performance of OCTA in discriminating the different stages of DR. Because many of the alterations in the moderate to severe stage of NPDR were found to be peripheral in the present study, wide-field OCT appears to be better suitable than a 3 × 3-mm2 or 6 × 6-mm2 field OCTA to detect areas of capillary loss.

    Our results were in agreement with several previous reports20,21 involving wide-field imaging modalities, including fluorescein angiography and OCTA. By stitching several fields together21 or implementing an extra lens to expand the scanning area,20 prevalent peripheral capillary dropout in DR was reported. Imaging far periphery vascularization is technically difficult. Using an ultrafast Multi-Megahertz OCT system (Optores GmbH) and special ultrawide-field ophthalmic optics design can potentially resolve the problem.30 We found that 12 × 12-mm2 wide-field OCTA had better diagnostic ability for NPDR than a central 6 × 6-mm2 cropped field. Recently, Hirano and colleagues31 compared the discriminative power of DR presence in different scanning protocols (3 × 3-mm2, 6 × 6-mm2, and 12 × 12-mm2 fields), and their results suggest that 3 × 3-mm2 field images might best determine DR. The discrepancy with the present study could be partially explained by inclusion of patients with proliferative DR for which neovascularization in the periphery may confound the quantification of OCTA-based vascular metrics. Moreover, results may depend on the exact parameters of image acquisition including A-scan rate, B-scan rate, and oversampling density as well as the filters and algorithms used for quantification of vascular metrics.32,33

    In the present study, we chose the approach of separating larger vessels from capillaries to assess the diagnostic performance of OCTA for DR. This approach was chosen because large vessels have a substantial contribution to the total peripapillary perfusion. Moreover, larger venular caliber and smaller arterial caliber are associated with DR progression and may therefore confound the association between capillary dropout and the stage of DR.34,35 Also, in the early stages of diabetes, retinal perfusion may be increased.36,37 In OCTA images, venules could be particularly important confounders for assessing DR because of their large diameters. Our results suggest that excluding the large vessels from perfusion density quantification provides extra sensitivity to NPDR diagnostic performance. A similar concept has been applied recently in calculating peripapillary perfused capillary density in patients with glaucoma38 and macula capillary density in patients with DR.39 Several studies40-42 have shown that subtraction of the peripapillary large vessels is associated with enhanced diagnostic accuracy of peripapillary OCTA for glaucoma. Moreover, Rosen et al39 detected early macular capillary changes in participants with diabetes without DR by removing the large vessels from the imaged area. By contrast, in our study, no improvement in diagnostic performance was found between the control group and the group with diabetes and no DR, most likely because of the good diabetes control and/or use of a wide-field approach in which macular vessel contributed little to the overall vascular metrics.

    Limitations

    This study has limitations. Our results are based on a relatively small sample size. Participants with severe NPDR were combined with participants with moderate NPDR because of insufficient numbers. In this study, we excluded patients who had undergone laser photocoagulation. The practice pattern in many centers in Singapore is to treat patients with severe NPDR with photocoagulation early before the progression to proliferative DR. Therefore, the prevalence of untreated severe NPDR in our population was low. During clinical recruitment, 9 of 17 participants (53%) with mild NPDR (25 eyes [32.9%]) and moderate to severe NPDR (28 eyes [36.8%]) had already received laser treatment and were excluded from this analysis. In the national DR screening program, similar to many other developed country programs, the definition for referable DR is moderate or worse NPDR. Therefore, the stratification of moderate or worse NPDR from mild NPDR or no DR still represents an important clinical outcome.

    The large vessel segmentation algorithm used in the present study was based on several vessel enhancement filters that assume a tubelike structure of the vessels. Segmentation errors with appearance of vessel branching, crossing, or vessel tortuosity were not evaluated systematically but appeared to be small. The ICC value was modest based on the small sample of older persons. The usability of the current algorithm needs to be further validated in a larger and more representative group of individuals. The initial quality check of wide-field OCTA excluded 30% to 40% of these scans because of multiple artifacts, such as low contrast, extensive motion artifacts, layer segmentation errors, and fixation problems. Exclusion of these scans might bring a bias to data selection because imaging may be more challenging for participants with worse visual acuity. The value of grading severity of NPDR is in the associated risk of progression to proliferative DR, which guides the management and follow-up of these patients.

    This is a retrospective case-control study that provides no insight into how well the OCTA capillary defect correlates with DR progression. Progression of DR and diabetic macular edema are associated with features identified on fundus fluorescein angiography,24 but the association with OCTA metrics is not well established because there are limited longitudinal data from this recently developed technique. One study showed that OCTA metrics, especially in deep capillary plexus, predicted DR progression43 based on a small field of view (3 × 3-mm2 field). An OCTA with higher speed and better tracking systems as used in the present study may enable wide-field OCTA scan to be captured in a single shot and thus may show better predictive value for DR progression.

    Conclusions

    Wide-field OCTA imaging may be useful for assessing peripheral capillary perfusion in eyes with DR. Isolation of large vessels from capillary perfusion density calculation may be associated with improved sensitivity to detect DR and may therefore be useful for disease prediction. The approach chosen in the present study may be particularly useful to detect peripheral capillary nonperfusion in patients with diabetes.

    Back to top
    Article Information

    Accepted for Publication: November 11, 2019.

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2020 Tan B et al. JAMA Network Open.

    Published: January 17, 2020. doi:10.1001/jamanetworkopen.2019.19469

    Correction: This article was corrected on June 5, 2020, to add grant numbers to the Funding/Support section.

    Corresponding Author: Gavin S. Tan, FRCSEd, PhD, Singapore Eye Research Institute, Singapore National Eye Centre, 20 College Rd, The Academia, Level 6, Discovery Tower, Singapore 169856 (gavin.tan@singhealth.com.sg).

    Author Contributions: Dr G. S. Tan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: T. Y. Wong, Schmetterer, G. S. Tan.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: B. Tan, Lin, Cheng, Gan, D. W. K. Wong, T. Y. Wong, Schmetterer, G. S. Tan.

    Critical revision of the manuscript for important intellectual content: B. Tan, Chua, Yao, Sabanayagam, D. Wong, Chan, T. Y. Wong, Schmetterer, G. S. Tan.

    Statistical analysis: Chua, Gan, D. W. K. Wong, Sabanayagam.

    Obtained funding: Schmetterer, G. S. Tan.

    Administrative, technical, or material support: Chua, Lin, Cheng, Chan, T. Y. Wong, Schmetterer, G. S. Tan.

    Supervision: Schmetterer, G. S. Tan.

    Conflict of Interest Disclosures: Dr D. Wong reported receiving lecture fees from Alcon and honorarium from Zeiss outside the submitted work. Dr G. S. Tan reported receiving grants from the Ministry of Health, National Medical Research Council during the conduct of the study; receiving travel support from Carl Zeiss Pte Ltd; receiving personal fees and consultancy fees from Novartis; receiving grants from Bayer AG and Santen Pharmaceutical Company, Ltd; and receiving travel support from Alcon outside the submitted work. No other disclosures were reported.

    Funding/Support: The study was funded by grants CG/C010A/2017, OFIRG/0048/2017, OFLCG/001/2017, TA/MOH-0055/2017, and TA/MOH-000249-00/2018 from the National Medical Research Council and grant Duke-NUS-KP(Coll)/2018/0009A from the Duke–National University of Singapore Medical School.

    Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    References
    1.
    Cheung  N, Mitchell  P, Wong  TY.  Diabetic retinopathy.   Lancet. 2010;376(9735):124-136. doi:10.1016/S0140-6736(09)62124-3 PubMedGoogle ScholarCrossref
    2.
    Wong  TY, Cheung  CMG, Larsen  M, Sharma  S, Simó  R.  Diabetic retinopathy.   Nat Rev Dis Primers. 2016;2:16012. doi:10.1038/nrdp.2016.12 PubMedGoogle ScholarCrossref
    3.
    Yau  JWY, Rogers  SL, Kawasaki  R,  et al; Meta-Analysis for Eye Disease (META-EYE) Study Group.  Global prevalence and major risk factors of diabetic retinopathy.   Diabetes Care. 2012;35(3):556-564. doi:10.2337/dc11-1909 PubMedGoogle ScholarCrossref
    4.
    Ting  DSW, Cheung  GCM, Wong  TY.  Diabetic retinopathy: global prevalence, major risk factors, screening practices and public health challenges: a review.   Clin Exp Ophthalmol. 2016;44(4):260-277. doi:10.1111/ceo.12696 PubMedGoogle ScholarCrossref
    5.
    Jia  Y, Bailey  ST, Hwang  TS,  et al.  Quantitative optical coherence tomography angiography of vascular abnormalities in the living human eye.   Proc Natl Acad Sci U S A. 2015;112(18):E2395-E2402. doi:10.1073/pnas.1500185112 PubMedGoogle ScholarCrossref
    6.
    Chua  J, Tan  B, Ang  M,  et al.  Future clinical applicability of optical coherence tomography angiography.   Clin Exp Optom. 2019;102(3):260-269. doi:10.1111/cxo.12854 PubMedGoogle ScholarCrossref
    7.
    Spaide  RF, Fujimoto  JG, Waheed  NK, Sadda  SR, Staurenghi  G.  Optical coherence tomography angiography.   Prog Retin Eye Res. 2018;64:1-55. doi:10.1016/j.preteyeres.2017.11.003 PubMedGoogle ScholarCrossref
    8.
    Ang  M, Tan  ACS, Cheung  CMG,  et al.  Optical coherence tomography angiography: a review of current and future clinical applications.   Graefes Arch Clin Exp Ophthalmol. 2018;256(2):237-245. doi:10.1007/s00417-017-3896-2 PubMedGoogle ScholarCrossref
    9.
    Dupas  B, Minvielle  W, Bonnin  S,  et al.  Association between vessel density and visual acuity in patients with diabetic retinopathy and poorly controlled type 1 diabetes.   JAMA Ophthalmol. 2018;136(7):721-728. doi:10.1001/jamaophthalmol.2018.1319 PubMedGoogle ScholarCrossref
    10.
    Durbin  MK, An  L, Shemonski  ND,  et al.  Quantification of retinal microvascular density in optical coherence tomographic angiography images in diabetic retinopathy.   JAMA Ophthalmol. 2017;135(4):370-376. doi:10.1001/jamaophthalmol.2017.0080 PubMedGoogle ScholarCrossref
    11.
    Agemy  SA, Scripsema  NK, Shah  CM,  et al.  Retinal vascular perfusion density mapping using optical coherence tomography angiography in normals and diabetic retinopathy patients.   Retina. 2015;35(11):2353-2363. doi:10.1097/IAE.0000000000000862 PubMedGoogle ScholarCrossref
    12.
    Fawzi  AA, Fayed  AE, Linsenmeier  RA, Gao  J, Yu  F.  Improved macular capillary flow on optical coherence tomography angiography after panretinal photocoagulation for proliferative diabetic retinopathy.   Am J Ophthalmol. 2019;206:217-227. doi:10.1016/j.ajo.2019.04.032 PubMedGoogle ScholarCrossref
    13.
    Hwang  TS, Gao  SS, Liu  L,  et al.  Automated quantification of capillary nonperfusion using optical coherence tomography angiography in diabetic retinopathy.   JAMA Ophthalmol. 2016;134(4):367-373. doi:10.1001/jamaophthalmol.2015.5658 PubMedGoogle ScholarCrossref
    14.
    Samara  WA, Shahlaee  A, Adam  MK,  et al.  Quantification of diabetic macular ischemia using optical coherence tomography angiography and its relationship with visual acuity.   Ophthalmology. 2017;124(2):235-244. doi:10.1016/j.ophtha.2016.10.008 PubMedGoogle ScholarCrossref
    15.
    Ishibazawa  A, Nagaoka  T, Takahashi  A,  et al.  Optical coherence tomography angiography in diabetic retinopathy: a prospective pilot study.   Am J Ophthalmol. 2015;160(1):35-44.e1. doi:10.1016/j.ajo.2015.04.021 PubMedGoogle ScholarCrossref
    16.
    Couturier  A, Mané  V, Bonnin  S,  et al.  Capillary plexus anomalies in diabetic retinopathy on optical coherence tomography angiography.   Retina. 2015;35(11):2384-2391. doi:10.1097/IAE.0000000000000859 PubMedGoogle ScholarCrossref
    17.
    Silva  PS, Dela Cruz  AJ, Ledesma  MG,  et al.  Diabetic retinopathy severity and peripheral lesions are associated with nonperfusion on ultrawide field angiography.   Ophthalmology. 2015;122(12):2465-2472. doi:10.1016/j.ophtha.2015.07.034 PubMedGoogle ScholarCrossref
    18.
    Silva  PS, Cavallerano  JD, Haddad  NMN,  et al.  Peripheral lesions identified on ultrawide field imaging predict increased risk of diabetic retinopathy progression over 4 years.   Ophthalmology. 2015;122(5):949-956. doi:10.1016/j.ophtha.2015.01.008 PubMedGoogle ScholarCrossref
    19.
    McNabb  RP, Grewal  DS, Mehta  R,  et al.  Wide field of view swept-source optical coherence tomography for peripheral retinal disease.   Br J Ophthalmol. 2016;100(10):1377-1382. doi:10.1136/bjophthalmol-2015-307480 PubMedGoogle ScholarCrossref
    20.
    Hirano  T, Kakihara  S, Toriyama  Y, Nittala  MG, Murata  T, Sadda  S.  Wide-field en face swept-source optical coherence tomography angiography using extended field imaging in diabetic retinopathy.   Br J Ophthalmol. 2018;102(9):1199-1203. doi:10.1136/bjophthalmol-2017-311358 PubMedGoogle ScholarCrossref
    21.
    Zhang  Q, Rezaei  KA, Saraf  SS, Chu  Z, Wang  F, Wang  RK.  Ultra-wide optical coherence tomography angiography in diabetic retinopathy.   Quant Imaging Med Surg. 2018;8(8):743-753. doi:10.21037/qims.2018.09.02 PubMedGoogle ScholarCrossref
    22.
    Pellegrini  M, Cozzi  M, Staurenghi  G, Corvi  F.  Comparison of wide field optical coherence tomography angiography with extended field imaging and fluorescein angiography in retinal vascular disorders.   PLoS One. 2019;14(4):e0214892. doi:10.1371/journal.pone.0214892 PubMedGoogle Scholar
    23.
    World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.   JAMA. 2013;310(20):2191-2194. doi:10.1001/jama.2013.281053PubMedGoogle ScholarCrossref
    24.
    Early Treatment Diabetic Retinopathy Study Research Group.  Fluorescein angiographic risk factors for progression of diabetic retinopathy. ETDRS report number 13.   Ophthalmology. 1991;98(5)(suppl):834-840. doi:10.1016/S0161-6420(13)38015-4 PubMedGoogle ScholarCrossref
    25.
    Hong  JM, Tan  B, Gupta  P,  et al.  Intrasession repeatability of quantitative metrics using widefield optical coherence tomography angiography (OCTA) in old community-dwelling Asian adults.   Invest Ophthalmol Vis Sci. 2019;60(9):3067.Google Scholar
    26.
    Wang  RK, An  L, Francis  P, Wilson  DJ.  Depth-resolved imaging of capillary networks in retina and choroid using ultrahigh sensitive optical microangiography.   Opt Lett. 2010;35(9):1467-1469. doi:10.1364/OL.35.001467 PubMedGoogle ScholarCrossref
    27.
    De Pretto  LR, Moult  EM, Alibhai  AY,  et al.  Controlling for artifacts in widefield optical coherence tomography angiography measurements of non-perfusion area.   Sci Rep. 2019;9(1):9096. doi:10.1038/s41598-019-43958-1 PubMedGoogle ScholarCrossref
    28.
    Koo  TK, Li  MY.  A guideline of selecting and reporting intraclass correlation coefficients for reliability research.   J Chiropr Med. 2016;15(2):155-163. doi:10.1016/j.jcm.2016.02.012 PubMedGoogle ScholarCrossref
    29.
    DeLong  ER, DeLong  DM, Clarke-Pearson  DL.  Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach.   Biometrics. 1988;44(3):837-845. doi:10.2307/2531595 PubMedGoogle ScholarCrossref
    30.
    Kolb  JP, Klein  T, Kufner  CL, Wieser  W, Neubauer  AS, Huber  R.  Ultra-widefield retinal MHz-OCT imaging with up to 100 degrees viewing angle.   Biomed Opt Express. 2015;6(5):1534-1552. doi:10.1364/BOE.6.001534 PubMedGoogle ScholarCrossref
    31.
    Hirano  T, Kitahara  J, Toriyama  Y, Kasamatsu  H, Murata  T, Sadda  S.  Quantifying vascular density and morphology using different swept-source optical coherence tomography angiographic scan patterns in diabetic retinopathy.   Br J Ophthalmol. 2019;103(2):216-221. doi:10.1136/bjophthalmol-2018-311942 PubMedGoogle ScholarCrossref
    32.
    Corvi  F, Pellegrini  M, Erba  S, Cozzi  M, Staurenghi  G, Giani  A.  Reproducibility of vessel density, fractal dimension, and foveal avascular zone using 7 different optical coherence tomography angiography devices.   Am J Ophthalmol. 2018;186:25-31. doi:10.1016/j.ajo.2017.11.011 PubMedGoogle ScholarCrossref
    33.
    Tan  B, Chua  J, Barathi  VA,  et al.  Quantitative analysis of choriocapillaris in non-human primates using swept-source optical coherence tomography angiography (SS-OCTA).   Biomed Opt Express. 2019;10(1):356-371. doi:10.1364/BOE.10.000356 PubMedGoogle Scholar
    34.
    Sun  C, Wang  JJ, Mackey  DA, Wong  TY.  Retinal vascular caliber: systemic, environmental, and genetic associations.   Surv Ophthalmol. 2009;54(1):74-95. doi:10.1016/j.survophthal.2008.10.003 PubMedGoogle ScholarCrossref
    35.
    Ikram  MK, Ong  YT, Cheung  CY, Wong  TY.  Retinal vascular caliber measurements: clinical significance, current knowledge and future perspectives.   Ophthalmologica. 2013;229(3):125-136. doi:10.1159/000342158 PubMedGoogle ScholarCrossref
    36.
    Pemp  B, Polska  E, Garhofer  G, Bayerle-Eder  M, Kautzky-Willer  A, Schmetterer  L.  Retinal blood flow in type 1 diabetic patients with no or mild diabetic retinopathy during euglycemic clamp.   Diabetes Care. 2010;33(9):2038-2042. doi:10.2337/dc10-0502 PubMedGoogle ScholarCrossref
    37.
    Fondi  K, Wozniak  PA, Howorka  K,  et al.  Retinal oxygen extraction in individuals with type 1 diabetes with no or mild diabetic retinopathy.   Diabetologia. 2017;60(8):1534-1540. doi:10.1007/s00125-017-4309-0 PubMedGoogle ScholarCrossref
    38.
    Suwan  Y, Fard  MA, Geyman  LS,  et al.  Association of myopia with peripapillary perfused capillary density in patients with glaucoma: an optical coherence tomography angiography study.   JAMA Ophthalmol. 2018;136(5):507-513. doi:10.1001/jamaophthalmol.2018.0776 PubMedGoogle ScholarCrossref
    39.
    Rosen  RB, Andrade Romo  JS, Krawitz  BD,  et al.  Earliest evidence of preclinical diabetic retinopathy revealed using optical coherence tomography angiography perfused capillary density.   Am J Ophthalmol. 2019;203:103-115. doi:10.1016/j.ajo.2019.01.012 PubMedGoogle ScholarCrossref
    40.
    Geyman  LS, Garg  RA, Suwan  Y,  et al.  Peripapillary perfused capillary density in primary open-angle glaucoma across disease stage: an optical coherence tomography angiography study.   Br J Ophthalmol. 2017;101(9):1261-1268. doi:10.1136/bjophthalmol-2016-309642 PubMedGoogle ScholarCrossref
    41.
    Liu  L, Jia  Y, Takusagawa  HL,  et al.  Optical coherence tomography angiography of the peripapillary retina in glaucoma.   JAMA Ophthalmol. 2015;133(9):1045-1052. doi:10.1001/jamaophthalmol.2015.2225 PubMedGoogle ScholarCrossref
    42.
    Rao  HL, Pradhan  ZS, Weinreb  RN,  et al.  Regional comparisons of optical coherence tomography angiography vessel density in primary open-angle glaucoma.   Am J Ophthalmol. 2016;171:75-83. doi:10.1016/j.ajo.2016.08.030 PubMedGoogle ScholarCrossref
    43.
    Sun  Z, Tang  FY, Wong  R,  et al.  OCT angiography metrics predict progression of diabetic retinopathy and development of diabetic macular edema: a prospective study.   Ophthalmology. 2019;126(12):1675-1684. doi:10.1016/j.ophtha.2019.06.016 PubMedGoogle Scholar
    ×