{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,7]],"date-time":"2026-10-07T04:52:09Z","timestamp":1791348729230,"version":"4.3.1"},"reference-count":13,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,10,7]],"date-time":"2026-10-07T00:00:00Z","timestamp":1791331200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Ophthalmol."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Cystoid macular edema (CME) is a vision-threatening complication of retinal vasculitis, a rare inflammatory disease for which large imaging datasets are difficult to obtain. Detection of CME on fluorescein angiography (FA) requires expert interpretation, motivating automated image-based approaches to support clinical decision-making. Deep learning methods, particularly convolutional neural networks (CNNs), have shown strong performance in medical image analysis but are constrained by limited labeled data in rare diseases. Transfer learning may mitigate this limitation by adapting pretrained models to small, task-specific datasets.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We analyzed 207 de-identified late-phase 55\u00b0 FA images from 117 patients with retinal vasculitis. The primary target was concurrent patient-level CME status, established through specialist review of clinical records and optical coherence tomography obtained at the same clinical visit as FA. A patient was classified as CME-positive when CME was present in at least one eye. An ImageNet-pretrained ResNet-18 was evaluated using five-fold nested, patient-level cross-validation. Within each outer-development cohort, 144 candidate configurations were compared by four-fold inner cross-validation and selected according to mean inner-validation AUROC. Each selected configuration was evaluated only on its corresponding untouched outer-test fold.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Across 117 pooled patient-level outer-test predictions, the model-selection pipeline achieved an AUROC of 0.708 (95% CI, 0.610\u20130.802), an AUPRC of 0.626 (95% CI, 0.493\u20130.755), and an accuracy of 0.726 (95% CI, 0.641\u20130.803). Sensitivity was 0.694 (95% CI, 0.558\u20130.822), specificity was 0.750 (95% CI, 0.643\u20130.853), and the F1 score was 0.680 (95% CI, 0.563\u20130.780). Exploratory Grad-CAM analysis visualized regions associated with model predictions.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>The nested model-selection pipeline showed moderate discrimination with a relatively balanced operating profile and slightly higher specificity than sensitivity. Clinically relevant false-positive and false-negative errors, together with the absence of external validation, preclude clinical deployment.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fopht.2026.1901237","type":"journal-article","created":{"date-parts":[[2026,10,7]],"date-time":"2026-10-07T04:48:47Z","timestamp":1791348527000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Classification of cystoid macular edema status from retinal vasculitis fluorescein angiography images using a convolutional neural network"],"prefix":"10.3389","volume":"6","author":[{"given":"Celina","family":"Wang","sequence":"first","affiliation":[{"name":"Milton Academy","place":["Milton, MA, United States"]},{"name":"Massachusetts Eye Research and Surgery Institution","place":["Waltham, MA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Koushani","family":"Chakrabarty","sequence":"additional","affiliation":[{"name":"Department of Biomedical Engineering, University at Buffalo","place":["Buffalo, NY, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Stephen D.","family":"Anesi","sequence":"additional","affiliation":[{"name":"Massachusetts Eye Research and Surgery Institution","place":["Waltham, MA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Peter Y.","family":"Chang","sequence":"additional","affiliation":[{"name":"Massachusetts Eye Research and Surgery Institution","place":["Waltham, MA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,10,7]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","DOI":"10.2147\/OPTH.S4033","article-title":"Cystoid macular edema","volume":"2","author":"Rotsos","year":"2008","journal-title":"Clin Ophthalmol"},{"key":"B2","doi-asserted-by":"publisher","first-page":"111","DOI":"10.1038\/s41746-024-01101-z","article-title":"FFA-GPT: An automated pipeline for fundus fluorescein angiography interpretation and question-answer","volume":"7","author":"Chen","year":"2024","journal-title":"NPJ Digital Med"},{"key":"B3","doi-asserted-by":"publisher","DOI":"10.1007\/s11517-025-03314-0","article-title":"A novel deep learning framework for retinal disease detection leveraging contextual and local features cues from retinal images","volume":"63","author":"Khan","year":"2025","journal-title":"Med Biol Eng Comput"},{"key":"B4","doi-asserted-by":"publisher","first-page":"227","DOI":"10.1186\/s12880-024-01406-1","article-title":"Optimized deep CNN for detection and classification of diabetic retinopathy and diabetic macular edema","volume":"24","author":"Thanikachalam","year":"2024","journal-title":"BMC Med Imaging"},{"key":"B5","doi-asserted-by":"publisher","first-page":"30321","DOI":"10.1038\/s41598-024-81127-1","article-title":"Evaluating deep learning models for classifying OCT images with limited data and noisy labels","volume":"14","author":"Miladinovi\u0107","year":"2024","journal-title":"Sci Rep"},{"key":"B6","doi-asserted-by":"publisher","DOI":"10.1080\/09273948.2024.2305185","article-title":"Using deep learning to segment retinal vascular leakage and occlusion in retinal vasculitis","volume":"32","author":"Dhirachaikulpanich","year":"2024","journal-title":"Ocular Immunol Inflammation"},{"key":"B7","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2016.2535302","article-title":"Convolutional neural networks for medical image analysis: Full training or fine tuning","volume":"35","author":"Tajbakhsh","year":"2016","journal-title":"IEEE Trans Med Imaging"},{"key":"B8","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2016.90","article-title":"Deep residual learning for image recognition","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)","author":"He","year":"2016"},{"key":"B9","doi-asserted-by":"publisher","first-page":"3763","DOI":"10.1038\/s41598-025-87171-9","article-title":"A deep learning based model for diabetic retinopathy grading","volume":"15","author":"Akhtar","year":"2025","journal-title":"Sci Rep"},{"key":"B10","doi-asserted-by":"publisher","first-page":"16975","DOI":"10.1038\/s41598-025-01496-z","article-title":"Enhancing medical explainability in deep learning for age-related macular degeneration diagnosis","volume":"15","author":"Shi","year":"2025","journal-title":"Sci Rep"},{"key":"B11","doi-asserted-by":"publisher","DOI":"10.1016\/j.cosrev.2019.100203","article-title":"Application of deep learning for retinal image analysis: A review","volume":"35","author":"Badar","year":"2020","journal-title":"Comput Sci Rev"},{"key":"B12","doi-asserted-by":"publisher","first-page":"36910","DOI":"10.1038\/s41598-025-20872-3","article-title":"Fine-tuned ResNet34 for efficient brain tumor classification","volume":"15","author":"Shahin","year":"2025","journal-title":"Sci Rep"},{"key":"B13","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-019-01228-7","article-title":"Grad-CAM: Visual explanations from deep networks via gradient-based localization","volume":"128","author":"Selvaraju","year":"2020","journal-title":"Int J Comput Vision"}],"container-title":["Frontiers in Ophthalmology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fopht.2026.1901237\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,7]],"date-time":"2026-10-07T04:48:56Z","timestamp":1791348536000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fopht.2026.1901237\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10,7]]},"references-count":13,"alternative-id":["10.3389\/fopht.2026.1901237"],"URL":"https:\/\/doi.org\/10.3389\/fopht.2026.1901237","relation":{},"ISSN":["2674-0826"],"issn-type":[{"value":"2674-0826","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,7]]},"article-number":"1901237"}}