{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T18:13:28Z","timestamp":1788286408043,"version":"build-2803163510"},"reference-count":30,"publisher":"BMJ","license":[{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"funder":[{"name":"NIH","award":["1R21-EY033480-01A1"],"award-info":[{"award-number":["1R21-EY033480-01A1"]}]}],"content-domain":{"domain":["bmj.com"],"crossmark-restriction":true},"short-container-title":["Br J Ophthalmol"],"accepted":{"date-parts":[[2026,8,18]]},"abstract":"<jats:sec>\n                    <jats:title>Purpose<\/jats:title>\n                    <jats:p>The purpose of this study is to develop a fully automated system for determining the optimal time to begin anti-vascular endothelial growth factor (anti-VEGF) treatment, based on retinal changes observed in optical coherence tomography (OCT) images and visual acuity (VA) scores.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>This retrospective study included two cohorts: patients with intermediate dry age-related macular degeneration (AMD) who did not progress to neovascular AMD (NVAMD) and those who progressed and received anti-VEGF treatment. Subjects had \u22653 consecutive visits. Two tasks were defined: (1) develop a deep learning (DL) model to identify the need for anti-VEGF treatment and (2) estimate outer retina thickness (ORT) from Bruch\u2019s membrane to the outer plexiform layer as an imaging biomarker. A convolutional neural network extracted OCT features related to ORT, and two long short-term memory models were trained on separate datasets to address Task I and II treatments.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      A total of 122\u2009patients (207 eyes) were included: 49 in dataset 1 (69 eyes; mean (SD) age 84 (8); 25 (51%) female) and 100 in dataset 2 (138 eyes; 82 (7); 54 (54%) female).\n                      <jats:italic>Task I<\/jats:italic>\n                      performance was evaluated by the model\u2019s alignment with clinical decisions on anti-VEGF initiation, and demonstrated area under the receiver operating curve of 0.73 (95% CI 0.59 to 0.87), area under the precision-recall curve of 0.51 (95% CI 0.35 to 0.66), an accuracy of 0.87 (95% CI 0.76 to 0.97), a precision of 0.73 (95% CI 0.59 to 0.87), a sensitivity of 0.69 (95% CI 0.54 to 0.83) and a specificity of 0.92 (95% CI 0.84 to 1.00).\n                      <jats:italic>Task II<\/jats:italic>\n                      achieved NMAE 0.12 (95% CI 0.04 to 0.20).\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>This DL framework provides a fully automated approach for determining optimal treatment timing, supporting earlier intervention to preserve VA in NVAMD patients.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Transitional relevance<\/jats:title>\n                    <jats:p>An accurate DL framework that can help with identifying the best time to initiate therapy for NVAMD.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1136\/bjo-2025-328327","type":"journal-article","created":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T17:20:19Z","timestamp":1788283219000},"page":"bjo-2025-328327","update-policy":"https:\/\/doi.org\/10.1136\/crossmarkpolicy","source":"Crossref","is-referenced-by-count":0,"title":["Modular multi-task deep learning framework for prediction of treatment initiation in neovascular age-related macular degeneration modular AI for NVAMD treatment initiation study-MANTIS"],"prefix":"10.1136","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-5542-3631","authenticated-orcid":false,"given":"Qitong","family":"Gao","sequence":"first","affiliation":[{"name":"Department of Electrical and Computer Engineering, 100 Science Dr, Durham, NC 27708, Duke University, Durham, North Carolina, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"David","family":"Kuo","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology, 2351 Erwin Rd, Durham, NC 27705, Duke University, Durham, North Carolina, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Joshua","family":"Amason","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology, 2351 Erwin Rd, Durham, NC 27705, Duke University, Durham, North Carolina, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Terry","family":"Lee","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology, 2351 Erwin Rd, Durham, NC 27705, Duke University, Durham, North Carolina, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jay K","family":"Rathinavelu","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology, 2351 Erwin Rd, Durham, NC 27705, Duke University, Durham, North Carolina, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Miroslav","family":"Pajic","sequence":"additional","affiliation":[{"name":"Department of Electrical and Computer Engineering, 100 Science Dr, Durham, NC 27708, Duke University, Durham, North Carolina, USA"},{"name":"Department of Computer Science, 308 Research Dr, Durham, NC 27705, Duke University, Durham, North Carolina, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4676-1844","authenticated-orcid":false,"given":"Majda","family":"Hadziahmetovic","sequence":"additional","affiliation":[{"name":"Department of Electrical and Computer Engineering, 100 Science Dr, Durham, NC 27708, Duke University, Durham, North Carolina, USA"},{"name":"Department of Ophthalmology, 2351 Erwin Rd, Durham, NC 27705, Duke University, Durham, North Carolina, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"239","published-online":{"date-parts":[[2026,9,1]]},"reference":[{"key":"2026090110201030000_bjo-2025-328327v1.1","doi-asserted-by":"publisher","DOI":"10.1167\/iovs.18-24106"},{"key":"2026090110201030000_bjo-2025-328327v1.2","doi-asserted-by":"publisher","DOI":"10.1007\/s00417-010-1520-9"},{"key":"2026090110201030000_bjo-2025-328327v1.3","doi-asserted-by":"publisher","DOI":"10.1167\/iovs.12-10361"},{"key":"2026090110201030000_bjo-2025-328327v1.4","doi-asserted-by":"crossref","DOI":"10.1038\/s41598-020-72359-y","article-title":"Prediction of age-related macular degeneration disease using a sequential deep learning approach on longitudinal SD-OCT imaging biomarkers","volume":"10","author":"Banerjee","year":"2020","journal-title":"Sci Rep"},{"key":"2026090110201030000_bjo-2025-328327v1.5","doi-asserted-by":"crossref","DOI":"10.1167\/tvst.9.2.25","article-title":"Artificial Intelligence to Stratify Severity of Age-Related Macular Degeneration (AMD) and Predict Risk of Progression to Late AMD","volume":"9","author":"Bhuiyan","year":"2020","journal-title":"Transl Vis Sci Technol"},{"key":"2026090110201030000_bjo-2025-328327v1.6","doi-asserted-by":"publisher","DOI":"10.1167\/iovs.14-14918"},{"key":"2026090110201030000_bjo-2025-328327v1.7","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pdig.0000106","article-title":"Automated large-scale prediction of exudative AMD progression using machine-read OCT biomarkers","volume":"2","author":"Rudas","year":"2023","journal-title":"PLOS Digit Health"},{"key":"2026090110201030000_bjo-2025-328327v1.8","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-020-0867-7"},{"key":"2026090110201030000_bjo-2025-328327v1.9","doi-asserted-by":"publisher","DOI":"10.1177\/15353702211031547"},{"key":"2026090110201030000_bjo-2025-328327v1.10","doi-asserted-by":"publisher","DOI":"10.1016\/j.ajo.2020.12.012"},{"key":"2026090110201030000_bjo-2025-328327v1.11","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1016\/j.ajo.2020.12.034","article-title":"Quantification of Key Retinal Features in Early and Late Age-Related Macular Degeneration Using Deep Learning","volume":"226","author":"Liefers","year":"2021","journal-title":"Am J Ophthalmol"},{"key":"2026090110201030000_bjo-2025-328327v1.12","doi-asserted-by":"crossref","first-page":"e693","DOI":"10.1016\/S2589-7500(21)00216-8","article-title":"Evaluation of artificial intelligence on a reference standard based on subjective interpretation","volume":"3","author":"Chen","year":"2021","journal-title":"Lancet Digit Health"},{"key":"2026090110201030000_bjo-2025-328327v1.13","doi-asserted-by":"crossref","DOI":"10.1016\/j.xops.2023.100421","article-title":"Impact of Anti-VEGF Treatment and Patient Characteristics on Vision Outcomes in Neovascular Age-related Macular Degeneration: Up to 6-Year Analysis of the AAO IRIS\u00ae Registry","volume":"4","author":"Wykoff","year":"2024","journal-title":"Ophthalmol Sci"},{"key":"2026090110201030000_bjo-2025-328327v1.14","doi-asserted-by":"crossref","first-page":"53","DOI":"10.1038\/eye.2016.260","article-title":"Impact of optical coherence tomography scanning density on quantitative analyses in neovascular age-related macular degeneration","volume":"31","author":"Velaga","year":"2017","journal-title":"Eye (Lond)"},{"key":"2026090110201030000_bjo-2025-328327v1.15","unstructured":"Enaholo ES , Musa MJ , Zeppieri M . Optical Coherence Tomography. StatPearls, 2025."},{"key":"2026090110201030000_bjo-2025-328327v1.16","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2018.02.010"},{"key":"2026090110201030000_bjo-2025-328327v1.17","doi-asserted-by":"crossref","unstructured":"Torrey L , Shavlik J . Transfer learning. Handbook of research on machine learning applications and trends: algorithms, methods, and techniques. IGI Global 2010;242\u201364. doi:10.4018\/978-1-60566-766-9.ch011","DOI":"10.4018\/978-1-60566-766-9.ch011"},{"key":"2026090110201030000_bjo-2025-328327v1.18","doi-asserted-by":"crossref","unstructured":"Szegedy C , Vanhoucke V , Ioffe S , et al . Rethinking the inception architecture for computer vision. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR); 2818\u201326. Las Vegas, NV, USA. doi:10.1109\/CVPR.2016.308","DOI":"10.1109\/CVPR.2016.308"},{"key":"2026090110201030000_bjo-2025-328327v1.19","first-page":"19290","article-title":"Rethinking the value of labels for improving class-imbalanced learning","volume":"33","author":"Yang","year":"2020","journal-title":"Adv Neural Inf Process Syst"},{"key":"2026090110201030000_bjo-2025-328327v1.20","doi-asserted-by":"crossref","first-page":"173","DOI":"10.4103\/sjopt.sjopt_91_23","article-title":"Vision transformers: The next frontier for deep learning-based ophthalmic image analysis","volume":"37","author":"Wu","year":"2023","journal-title":"Saudi J Ophthalmol"},{"key":"2026090110201030000_bjo-2025-328327v1.21","doi-asserted-by":"crossref","DOI":"10.1038\/s41598-025-02299-y","article-title":"Residual self-attention vision transformer for detecting acquired vitelliform lesions and age-related macular drusen","volume":"15","author":"Powroznik","year":"2025","journal-title":"Sci Rep"},{"key":"2026090110201030000_bjo-2025-328327v1.22","doi-asserted-by":"publisher","DOI":"10.1148\/radiology.143.1.7063747"},{"key":"2026090110201030000_bjo-2025-328327v1.23","doi-asserted-by":"publisher","DOI":"10.2307\/2276774"},{"key":"2026090110201030000_bjo-2025-328327v1.24","doi-asserted-by":"publisher","DOI":"10.1111\/opo.12675"},{"key":"2026090110201030000_bjo-2025-328327v1.25","doi-asserted-by":"crossref","first-page":"141","DOI":"10.1038\/s42256-020-0154-9","article-title":"Deep-learning-based Prediction of Late Age-Related Macular Degeneration Progression","volume":"2","author":"Yan","year":"2020","journal-title":"Nat Mach Intell"},{"key":"2026090110201030000_bjo-2025-328327v1.26","unstructured":"Pfeiffer J , Ruder S , Vuli\u0107 I , et al . Modular Deep Learning. Transactions on Machine Learning Research 2023."},{"key":"2026090110201030000_bjo-2025-328327v1.27","unstructured":"Ponti EM , Sordoni A , Bengio Y , et al . Combining modular skills in multitask learning. arXiv 2022."},{"key":"2026090110201030000_bjo-2025-328327v1.28","doi-asserted-by":"crossref","unstructured":"Selvaraju RR , Cogswell M , Das A , et al . Grad-cam: visual explanations from deep networks via gradient-based localization. 2017 IEEE International Conference on Computer Vision (ICCV); 618\u201326. Venice. doi:10.1109\/ICCV.2017.74","DOI":"10.1109\/ICCV.2017.74"},{"key":"2026090110201030000_bjo-2025-328327v1.29","doi-asserted-by":"publisher","DOI":"10.1016\/j.ajo.2011.03.008"},{"key":"2026090110201030000_bjo-2025-328327v1.30","doi-asserted-by":"crossref","first-page":"31","DOI":"10.1016\/j.oret.2019.08.003","article-title":"Prognostic Value of Retinal Layers in Comparison with Other Risk Factors for Conversion of Intermediate Age-related Macular Degeneration","volume":"4","author":"Thiele","year":"2020","journal-title":"Ophthalmol Retina"}],"container-title":["British Journal of Ophthalmology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1136\/bjo-2025-328327","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T17:20:22Z","timestamp":1788283222000},"score":1,"resource":{"primary":{"URL":"https:\/\/bjo.bmj.com\/lookup\/doi\/10.1136\/bjo-2025-328327"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,1]]},"references-count":30,"alternative-id":["10.1136\/bjo-2025-328327"],"URL":"https:\/\/doi.org\/10.1136\/bjo-2025-328327","relation":{},"ISSN":["0007-1161","1468-2079"],"issn-type":[{"value":"0007-1161","type":"print"},{"value":"1468-2079","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,1]]}}}