Abstract
Objectives
To demonstrate the feasibility of a deep learning-based vascular segmentation tool for UWFA and evaluate its ability to automatically identify quality-optimized phase-specific images.
Methods
Cumulative retinal vessel areas (RVA) were extracted from all available UWFA frames. Cubic splines were fitted for serial vascular assessment throughout the angiographic phases of eyes with diabetic retinopathy (DR), sickle cell retinopathy (SCR), or normal retinal vasculature. The image with maximum RVA was selected as the optimum early phase. A late phase frame was selected at a minimum of 4 min that most closely mirrored the RVA from the selected early image. Trained image analysts evaluated the selected pairs.
Results
A total of 13,980 UWFA sequences from 462 sessions were used to evaluate the performance and 1578 UWFA sequences from 66 sessions were used to create cubic splines. Maximum RVA was detected at a mean of 41 ± 15, 47 ± 27, 38 ± 8 s for DR, SCR, and normals respectively. In 85.2% of the sessions, appropriate images for both phases were successfully identified. The individual success rate was 90.7% for early and 94.6% for late frames.
Conclusions
Retinal vascular characteristics are highly phased and field-of-view sensitive. Vascular parameters extracted by deep learning algorithms can be used for quality assessment of angiographic images and quality optimized phase selection. Clinical applications of a deep learning-based vascular segmentation and phase selection system might significantly improve the speed, consistency, and objectivity of UWFA evaluation.
Similar content being viewed by others
Log in or create a free account to read this content
Gain free access to this article, as well as selected content from this journal and more on nature.com
or
References
Cabrera DeBuc D, Somfai GM, Koller A. Retinal microvascular network alterations: potential biomarkers of cerebrovascular and neural diseases. Am J Physiol Heart Circ Physiol. 2017;312:H201–H12.
MacGillivray TJ, Trucco E, Cameron JR, Dhillon B, Houston JG, van Beek EJ. Retinal imaging as a source of biomarkers for diagnosis, characterization and prognosis of chronic illness or long-term conditions. Br J Radio. 2014;87:20130832.
Poplin R, Varadarajan AV, Blumer K, Liu Y, McConnell MV, Corrado GS, et al. Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning. Nat Biomed Eng. 2018;2:158–64.
Liskowski P, Krawiec K. Segmenting retinal blood vessels with deep neural networks. IEEE Trans Med Imaging. 2016;35:2369–80.
Joonyoung S, Boreom L. Development of automatic retinal vessel segmentation method in fundus images via convolutional neural networks. Conf Proc IEEE Eng Med Biol Soc. 2017;2017:681–4.
Yin B, Li H, Sheng B, Hou X, Chen Y, Wu W, et al. Vessel extraction from non-fluorescein fundus images using orientation-aware detector. Med Image Anal. 2015;26:232–42.
Manivannan A, Plskova J, Farrow A, McKay S, Sharp PF, Forrester JV. Ultra-wide-field fluorescein angiography of the ocular fundus. Am J Ophthalmol. 2005;140:525–7.
Ghasemi Falavarjani K, Wang K, Khadamy J, Sadda SR. Ultra-wide-field imaging in diabetic retinopathy; an overview. J Curr Ophthalmol. 2016;28:57–60.
Ehlers JP, Jiang AC, Boss JD, Hu M, Figueiredo N, Babiuch A, et al. Quantitative ultra-widefield angiography and diabetic retinopathy severity: an assessment of panretinal leakage index, ischemic index and microaneurysm count. Ophthalmology 2019;126:1527–32.
Ding L, Kuriyan A, Ramchandran R, Sharma G. Quantification of longitudinal changes in retinal vasculature from wide-field flourescein angiography via a novel registration and change detection approach. IEEE international conference on acoustics, speech and signal processing (ICASSP). 2018;1070–4.
Ding L, Kuriyan A, Ramchandran R, Sharma G. MULTI-Scale morphological analysis for retinal vessel detection in wide-field flourescein angiography. IEEE Western New York Image and Signal Processing Workshop (WNYISPW). 2017;1–5
Fan W, Uji A, Borrelli E, Singer M, Sagong M, van Hemert J, et al. Precise measurement of retinal vascular bed area and density on ultra-wide fluorescein angiography in normal subjects. Am J Ophthalmol. 2018;188:155–63.
Moosavi A, Figueiredo N, Prasanna P, Srivastava SK, Sharma K, Madabhushi A, et al. Imaging features of vessels and leakage patterns predict extended interval aflibercept dosing using ultra-widefield angiography in retinal vascular disease: findings from the PERMEATE study. IEEE Trans Biomed Eng. 2021:68:1777–86.
Jiang A, Srivastava S, Figueiredo N, Babiuch A, Hu M, Reese J, et al. Repeatability of automated leakage quantification and microaneurysm identification utilising an analysis platform for ultra-widefield fluorescein angiography. Br J Ophthalmol. 2020;104:500–3.
Ehlers JP, Wang K, Vasanji A, Hu M, Srivastava SK. Automated quantitative characterisation of retinal vascular leakage and microaneurysms in ultra-widefield fluorescein angiography. Br J Ophthalmol. 2017;101:696–9.
U-Net: convolutional networks for biomedical image segmentation. 2015. https://arxiv.org/abs/1505.04597.
Peng C, Zhang X, Yu G, Luo G, Sun J. Large kernel matters—improve semantic segmentation by global convolutional network. 2017. https://arxiv.org/abs/1703.02719.
Croft DE, van Hemert J, Wykoff CC, Clifton D, Verhoek M, Fleming A. et al. Precise montaging and metric quantification of retinal surface area from ultra-widefield fundus photography and fluorescein angiography. Ophthalmic Surg Lasers Imaging Retina. 2014;45:312–7.
Funding
NIH/NEI K23-EY022947-01A1 (JPE), Betty J. Powers Retina Research Fellowship (DDS).
Author information
Authors and Affiliations
Contributions
DDS: performed data analysis, image analysis, and manuscript preparation/revisions; SKS: provided supervision, provided imaging data, manuscript revisions; CW: provided data acquisition and imaging data, manuscript revisions; AWS: provided data acquisition and imaging data, manuscript revisions; JH: performed image analysis, provided data organization, manuscript revisions; MO: performed image analysis, provided data organization, manuscript revisions; JW: provided segmentation expertise and analysis support., manuscript revisions; AV: provided segmentation expertise and analysis support, manuscript revisions; JLR: provided supervision, data resources, and manuscript revisions; JPE: provided project oversight, funding support, resource support, study planning, manuscript revisions.
Corresponding author
Ethics declarations
Competing interests
SKS receives funding from Gilead, Regeneron, and Allergan; receives compensation as a consultant from Bausch and Lomb and Santen; owns a patent with Leica. CW receives compensation as a consultant from Adverum, Allergan, Apellis, Clearside, EyePoint, Genentech/Roch, Neurotech, Novartis, Opthea, Regeneron, Regenxbio, Samsung, Santen, Alimera Sciences, Allegro, Alynylam, Bayer, Clearside, D.O.R.C., Kodiak, Notal Vision, ONL Therapeutics, PolyPhotonix, and RecensMedical. AWS receives compensation as a consultant from Allergan and Novartis. AV is an employee of ERT. JPE receives funding and compensation as a consultant from Aerpio, Adverum Alcon, Thrombogenics/Oxurion, Regeneron, Stealth, Roche, Genetech, Novartis, and Allergan; receives compensation as a consultant from Roche, Leica, Zeiss, Allegro, Santen and has a patent with Leica.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Sevgi, D.D., Srivastava, S.K., Wykoff, C. et al. Deep learning-enabled ultra-widefield retinal vessel segmentation with an automated quality-optimized angiographic phase selection tool. Eye 36, 1783–1788 (2022). https://doi.org/10.1038/s41433-021-01661-4
Received:
Revised:
Accepted:
Published:
Version of record:
Issue date:
DOI: https://doi.org/10.1038/s41433-021-01661-4
This article is cited by
-
Development and evaluation of a deep learning model for automatic segmentation of non-perfusion area in fundus fluorescein angiography
Journal of Big Data (2024)
-
Retinal non-perfusion in diabetic retinopathy
Eye (2022)
-
Deep learning for ultra-widefield imaging: a scoping review
Graefe's Archive for Clinical and Experimental Ophthalmology (2022)
