{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,29]],"date-time":"2026-09-29T05:00:00Z","timestamp":1790658000390,"version":"4.1.0"},"reference-count":42,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,9,29]],"date-time":"2026-09-29T00:00:00Z","timestamp":1790640000000},"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. Med."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Diabetic retinopathy (DR) is a leading cause of vision impairment worldwide, highlighting the need for accurate and accessible screening tools. Optical Coherence Tomography (OCT) provides high-resolution structural information of the retina, whereas OCT angiography (OCTA) offers complementary vascular information that is highly relevant for DR diagnosis.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>In this study, we propose a cross-modal fusion of OCT B-scans with single-channel enface OCTA using a bidirectional cross-modal attention network for automated DR classification. Two independent datasets, OCT500 and UIC, comprising 730 scans in total, were utilized to evaluate performance in within-dataset, combined-dataset, and cross-dataset generalization settings. A ConvNeXt V2 Large model trained solely on OCT images served as the unimodal baseline. In addition to ground-truth (GT) OCTA, we explored the use of translated (TR) OCTA generated from OCT scans, eliminating the requirement for dedicated OCTA hardware.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Experimental results demonstrated that cross-modal fusion consistently outperformed unimodal OCT classification across all evaluation scenarios. Fusion with GT OCTA improved classification accuracy and discriminative performance, while TR OCTA achieved comparable or superior results in most settings. Furthermore, TR OCTA improved sensitivity and cross-dataset generalization, indicating enhanced robustness to domain shifts.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>These findings demonstrate that attention-based OCTOCTA enface fusion provides clinically meaningful improvements for DR detection and suggest that computationally generated OCTA can serve as a practical, low-cost alternative to hardware-acquired OCTA, enabling broader deployment of high-performance retinal screening systems in resource-limited clinical environments.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fmed.2026.1900570","type":"journal-article","created":{"date-parts":[[2026,9,29]],"date-time":"2026-09-29T04:46:58Z","timestamp":1790657218000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Cross-modal fusion of OCT and OCT angiography enface for improved diagnostics of diabetic retinopathy"],"prefix":"10.3389","volume":"13","author":[{"given":"Rashadul Hasan","family":"Badhon","sequence":"first","affiliation":[{"name":"Department of Electrical and Computer Engineering, University of North Carolina at Charlotte","place":["Charlotte, NC, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Atalie Carina","family":"Thompson","sequence":"additional","affiliation":[{"name":"Department of Surgical Ophthalmology, Atrium-Health Wake Forest Baptist","place":["Winston-Salem, NC, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jennifer I.","family":"Lim","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology and Visual Science, University of Illinois at Chicago","place":["Chicago, IL, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Theodore","family":"Leng","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology, Stanford University School of Medicine","place":["Stanford, CA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Minhaj","family":"Nur Alam","sequence":"additional","affiliation":[{"name":"Department of Electrical and Computer Engineering, University of North Carolina at Charlotte","place":["Charlotte, NC, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,9,29]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"1178","DOI":"10.1126\/science.1957169","article-title":"Optical coherence tomography","volume":"254","author":"Huang","year":"1991","journal-title":"Science"},{"key":"B2","doi-asserted-by":"publisher","first-page":"239","DOI":"10.1088\/0034-4885\/66\/2\/204","article-title":"Optical coherence tomography\u2014principles and applications","volume":"66","author":"Fercher","year":"2003","journal-title":"Rep Prog Phys"},{"key":"B3","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-540-77550-8","volume-title":"Optical Coherence Tomography: Technology and Applications","author":"Drexler","year":"2008"},{"key":"B4","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.preteyeres.2018.07.004","article-title":"Artificial intelligence in retina","volume":"67","author":"Schmidt-Erfurth","year":"2018","journal-title":"Prog Retin Eye Res"},{"key":"B5","doi-asserted-by":"publisher","first-page":"066008","DOI":"10.1117\/1.JBO.22.6.066008","article-title":"Optical coherence tomography angiography-based capillary velocimetry","volume":"22","author":"Wang","year":"2017","journal-title":"J Biomed Opt"},{"key":"B6","doi-asserted-by":"publisher","first-page":"066008","DOI":"10.1117\/1.JBO.21.6.066008","article-title":"Quantitative assessment of the retinal microvasculature using optical coherence tomography angiography","volume":"21","author":"Chu","year":"2016","journal-title":"J Biomed Opt"},{"key":"B7","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.preteyeres.2017.11.003","article-title":"Optical coherence tomography angiography","volume":"64","author":"Spaide","year":"2018","journal-title":"Prog Retin Eye Res"},{"key":"B8","doi-asserted-by":"publisher","first-page":"13","DOI":"10.1016\/j.exer.2018.02.006","article-title":"Quantitative comparisons between optical coherence tomography angiography and matched histology in the human eye","volume":"170","author":"An","year":"2018","journal-title":"Exp Eye Res"},{"key":"B9","doi-asserted-by":"publisher","first-page":"3257","DOI":"10.1364\/BOE.10.003257","article-title":"Development and validation of a deep learning algorithm for distinguishing the nonperfusion area from signal reduction artifacts on OCT angiography","volume":"10","author":"Guo","year":"2019","journal-title":"Biomed Opt Express"},{"key":"B10","doi-asserted-by":"publisher","first-page":"15","DOI":"10.1038\/s42003-019-0730-x","article-title":"Deep learning for detecting retinal detachment and discerning macular status using ultra-widefield fundus images","volume":"3","author":"Li","year":"2020","journal-title":"Commun Biol"},{"key":"B11","doi-asserted-by":"publisher","first-page":"3568","DOI":"10.1364\/BOE.5.003568","article-title":"Fully automated detection of diabetic macular edema and dry age-related macular degeneration from optical coherence tomography images","volume":"5","author":"Srinivasan","year":"2014","journal-title":"Biomed Opt Express"},{"key":"B12","doi-asserted-by":"publisher","first-page":"1655453","DOI":"10.3389\/fmed.2025.1655453","article-title":"Diffusion model based OCT to OCTA translation","volume":"12","author":"Badhon","year":"2025","journal-title":"Front Med"},{"key":"B13","doi-asserted-by":"publisher","first-page":"e28868","DOI":"10.2196\/28868","article-title":"A multimodal imaging-based deep learning model for detecting treatment-requiring retinal vascular diseases: model development and validation study","volume":"9","author":"Kang","year":"2021","journal-title":"JMIR Med Inform"},{"key":"B14","first-page":"13594","article-title":"Diabetic retinopathy segmentation and classification using deep learning approach","volume":"25","author":"Nallasivan","year":"2021","journal-title":"Ann RSCB"},{"key":"B15","doi-asserted-by":"publisher","first-page":"102612","DOI":"10.1016\/j.media.2022.102612","article-title":"Cross-modal attention for multi-modal image registration","volume":"82","author":"Song","year":"2022","journal-title":"Med Image Anal"},{"key":"B16","doi-asserted-by":"publisher","first-page":"108635","DOI":"10.1016\/j.compbiomed.2024.108635","article-title":"A review of deep learning-based information fusion techniques for multimodal medical image classification","volume":"177","author":"Li","year":"2024","journal-title":"Comput Biol Med"},{"key":"B17","doi-asserted-by":"publisher","first-page":"105253","DOI":"10.1016\/j.compbiomed.2022.105253","article-title":"A review on multimodal medical image fusion: compendious analysis of medical modalities, multimodal databases, fusion techniques and quality metrics","volume":"144","author":"Azam","year":"2022","journal-title":"Comput Biol Med"},{"key":"B18","doi-asserted-by":"publisher","first-page":"673","DOI":"10.2174\/1573405618666220606161137","article-title":"Recent advancements in multimodal medical image fusion techniques for better diagnosis: an overview","volume":"19","author":"Haribabu","year":"2023","journal-title":"Curr Med Imaging"},{"key":"B19","doi-asserted-by":"publisher","first-page":"44","DOI":"10.1007\/s44196-023-00225-6","article-title":"A review of the application of multi-modal deep learning in medicine: bibliometrics and future directions","volume":"16","author":"Pei","year":"2023","journal-title":"Int J Comput Intell Syst"},{"key":"B20","doi-asserted-by":"publisher","first-page":"4713","DOI":"10.1364\/BOE.495999","article-title":"Optimizing the OCTA layer fusion option for deep learning classification of diabetic retinopathy","volume":"14","author":"Ebrahimi","year":"2023","journal-title":"Biomed Opt Express"},{"key":"B21","doi-asserted-by":"publisher","first-page":"20","DOI":"10.1167\/tvst.9.2.20","article-title":"Ensemble deep learning for diabetic retinopathy detection using optical coherence tomography angiography","volume":"9","author":"Heisler","year":"2020","journal-title":"Transl Vis Sci Technol"},{"key":"B22","doi-asserted-by":"publisher","first-page":"1591","DOI":"10.1109\/TMI.2021.3059956","article-title":"Multi-modal retinal image classification with modality-specific attention network","volume":"40","author":"He","year":"2021","journal-title":"IEEE Trans Med Imaging"},{"key":"B23","doi-asserted-by":"publisher","first-page":"4579","DOI":"10.21037\/qims-24-9","article-title":"CMAF-Net: a cross-modal attention fusion-based deep neural network for incomplete multi-modal brain tumor segmentation","volume":"14","author":"Sun","year":"2024","journal-title":"Quant Imaging Med Surg"},{"key":"B24","doi-asserted-by":"publisher","first-page":"102147","DOI":"10.1016\/j.inffus.2023.102147","article-title":"CrossFuse: a novel cross attention mechanism based infrared and visible image fusion approach","volume":"103","author":"Li","year":"2024","journal-title":"Inf Fusion"},{"key":"B25","doi-asserted-by":"publisher","first-page":"2126","DOI":"10.1609\/aaai.v36i2.20109","article-title":"TransMEF: a transformer-based multi-exposure image fusion framework using self-supervised multi-task learning","volume":"36","author":"Qu","year":"2022","journal-title":"Proc AAAI Conf Artif Intell"},{"key":"B26","doi-asserted-by":"publisher","first-page":"5134","DOI":"10.1109\/TIP.2022.3193288","article-title":"MATR: multimodal medical image fusion via multiscale adaptive transformer","volume":"31","author":"Tang","year":"2022","journal-title":"IEEE Trans Image Process"},{"key":"B27","doi-asserted-by":"publisher","first-page":"3317","DOI":"10.1109\/JBHI.2024.3391620","article-title":"MACTFusion: lightweight cross transformer for adaptive multimodal medical image fusion","volume":"29","author":"Xie","year":"2025","journal-title":"IEEE J Biomed Health Inform"},{"key":"B28","doi-asserted-by":"publisher","first-page":"11657","DOI":"10.1038\/s41598-025-93616-y","article-title":"A novel multimodal medical image fusion framework with edge enhancement and cross-scale transformer","volume":"15","author":"Luo","year":"2025","journal-title":"Sci Rep"},{"key":"B29","doi-asserted-by":"publisher","first-page":"e26783","DOI":"10.1002\/hbm.26783","article-title":"A multimodal vision transformer for interpretable fusion of functional and structural neuroimaging data","volume":"45","author":"Bi","year":"2024","journal-title":"Hum Brain Mapp"},{"key":"B30","doi-asserted-by":"publisher","first-page":"4155","DOI":"10.32604\/cmc.2025.065571","article-title":"A comprehensive review of multimodal deep learning for enhanced medical diagnostics","volume":"84","author":"Al-Zoghby","year":"2025","journal-title":"Comput Mater Contin"},{"key":"B31","doi-asserted-by":"publisher","first-page":"5249","DOI":"10.1364\/BOE.399514","article-title":"AV-Net: deep learning for fully automated artery-vein classification in optical coherence tomography angiography","volume":"11","author":"Alam","year":"2020","journal-title":"Biomed Opt Express"},{"key":"B32","doi-asserted-by":"publisher","first-page":"4870","DOI":"10.1364\/BOE.468483","article-title":"MF-AV-Net: an open-source deep learning network with multimodal fusion options for artery-vein segmentation in OCT angiography","volume":"13","author":"Abtahi","year":"2022","journal-title":"Biomed Opt Express"},{"key":"B33","doi-asserted-by":"publisher","first-page":"2770","DOI":"10.3390\/diagnostics13172770","article-title":"Hybrid fusion of high-resolution and ultra-widefield OCTA acquisitions for the automatic diagnosis of diabetic retinopathy","volume":"13","author":"Li","year":"2023","journal-title":"Diagnostics"},{"key":"B34","doi-asserted-by":"publisher","first-page":"045030","DOI":"10.1088\/2057-1976\/adeb92","article-title":"Multi-modal classification of retinal disease based on convolutional neural network","volume":"11","author":"Pan","year":"2025","journal-title":"Biomed Phys Eng Express"},{"key":"B35","doi-asserted-by":"publisher","first-page":"4316507","DOI":"10.1155\/2022\/4316507","article-title":"Diagnosing diabetic retinopathy in OCTA images based on multilevel information fusion using a deep learning framework","volume":"2022","author":"Li","year":"2022","journal-title":"Comput Math Methods Med"},{"key":"B36","doi-asserted-by":"publisher","first-page":"10","DOI":"10.1167\/tvst.11.7.10","article-title":"A diabetic retinopathy classification framework based on deep-learning analysis of OCT angiography","volume":"11","author":"Zang","year":"2022","journal-title":"Transl Vis Sci Technol"},{"key":"B37","doi-asserted-by":"publisher","first-page":"23024","DOI":"10.1038\/s41598-021-02479-6","article-title":"A deep learning model for identifying diabetic retinopathy using optical coherence tomography angiography","volume":"11","author":"Ryu","year":"2021","journal-title":"Sci Rep"},{"key":"B38","doi-asserted-by":"publisher","first-page":"bjo-2025-329249","DOI":"10.1136\/bjo-2025-329249","article-title":"Classification of retinal diseases based on optical coherence tomography angiography using cross-modal transfer learning of domain-specific foundation AI models","author":"Shah","year":"2026","journal-title":"Br J Ophthalmol"},{"key":"B39","doi-asserted-by":"publisher","first-page":"1711","DOI":"10.1109\/TMI.2024.3518067","article-title":"MultiEYE: dataset and benchmark for OCT-enhanced retinal disease recognition from fundus images","volume":"44","author":"Liu","year":"2025","journal-title":"IEEE Trans Med Imaging"},{"key":"B40","doi-asserted-by":"publisher","first-page":"103092","DOI":"10.1016\/j.media.2024.103092","article-title":"OCTA-500: a retinal dataset for optical coherence tomography angiography study","volume":"93","author":"Li","year":"2024","journal-title":"Med Image Anal"},{"key":"B41","author":"Woo","year":""},{"key":"B42","doi-asserted-by":"publisher","first-page":"3699","DOI":"10.1364\/BOE.516764","article-title":"Cross-modal attention network for retinal disease classification based on multi-modal images","volume":"15","author":"Liu","year":"2024","journal-title":"Biomed Opt Express"}],"container-title":["Frontiers in Medicine"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2026.1900570\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,29]],"date-time":"2026-09-29T04:47:03Z","timestamp":1790657223000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2026.1900570\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,29]]},"references-count":42,"alternative-id":["10.3389\/fmed.2026.1900570"],"URL":"https:\/\/doi.org\/10.3389\/fmed.2026.1900570","relation":{},"ISSN":["2296-858X"],"issn-type":[{"value":"2296-858X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,29]]},"article-number":"1900570"}}