{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,26]],"date-time":"2026-06-26T02:19:19Z","timestamp":1782440359892,"version":"3.54.5"},"reference-count":35,"publisher":"BMJ","issue":"5","license":[{"start":{"date-parts":[[2025,5,30]],"date-time":"2025-05-30T00:00:00Z","timestamp":1748563200000},"content-version":"unspecified","delay-in-days":29,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100021171","name":"Basic and Applied Basic Research Foundation of Guangdong Province","doi-asserted-by":"publisher","award":["210719096902710"],"award-info":[{"award-number":["210719096902710"]}],"id":[{"id":"10.13039\/501100021171","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Guangdong Province Science and Technology Special Fund","award":["210719096902710"],"award-info":[{"award-number":["210719096902710"]}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62206164"],"award-info":[{"award-number":["62206164"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["bmj.com"],"crossmark-restriction":true},"short-container-title":["BMJ Open"],"accepted":{"date-parts":[[2025,5,14]]},"published-print":{"date-parts":[[2025,5]]},"abstract":"<jats:sec><jats:title>Objectives<\/jats:title><jats:p>To develop and validate an automated diabetic macular oedema (DME) classification system based on the images from different three-dimensional optical coherence tomography (3-D OCT) devices.<\/jats:p><\/jats:sec><jats:sec><jats:title>Design<\/jats:title><jats:p>A multicentre, platform-based development study using retrospective and cross-sectional data. Data were subjected to a two-level grading system by trained graders and a retina specialist, and categorised into three types: no DME, non-centre-involved DME and centre-involved DME (CI-DME). The 3-D convolutional neural networks algorithm was used for DME classification system development. The deep learning (DL) performance was compared with the diabetic retinopathy experts.<\/jats:p><\/jats:sec><jats:sec><jats:title>Setting<\/jats:title><jats:p>Data were collected from Joint Shantou International Eye Center of Shantou University and the Chinese University of Hong Kong, Chaozhou People\u2019s Hospital and The Second Affiliated Hospital of Shantou University Medical College from January 2010 to December 2023.<\/jats:p><\/jats:sec><jats:sec><jats:title>Participants<\/jats:title><jats:p>7790 volumes of 7146 eyes from 4254 patients were annotated, of which 6281 images were used as the development set and 1509 images were used as the external validation set, split based on the centres.<\/jats:p><\/jats:sec><jats:sec><jats:title>Main outcomes<\/jats:title><jats:p>Accuracy, F1-score, sensitivity, specificity, area under receiver operating characteristic curve (AUROC) and Cohen\u2019s kappa were calculated to evaluate the performance of the DL algorithm.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>In classifying DME with non-DME, our model achieved an AUROCs of 0.990 (95% CI 0.983 to 0.996) and 0.916 (95% CI 0.902 to 0.930) for hold-out testing dataset and external validation dataset, respectively. To distinguish CI-DME from non-centre-involved-DME, our model achieved AUROCs of 0.859 (95% CI 0.812 to 0.906) and 0.881 (95% CI 0.859 to 0.902), respectively. In addition, our system showed comparable performance (Cohen\u2019s \u03ba: 0.85 and 0.75) to the retina experts (Cohen\u2019s \u03ba: 0.58\u20130.92 and 0.70\u20130.71).<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusions<\/jats:title><jats:p>Our DL system achieved high accuracy in multiclassification tasks on DME classification with 3-D OCT images, which can be applied to population-based DME screening.<\/jats:p><\/jats:sec>","DOI":"10.1136\/bmjopen-2025-099167","type":"journal-article","created":{"date-parts":[[2025,5,31]],"date-time":"2025-05-31T13:20:18Z","timestamp":1748697618000},"page":"e099167","update-policy":"https:\/\/doi.org\/10.1136\/crossmarkpolicy","source":"Crossref","is-referenced-by-count":5,"title":["Development and validation of a 3-D deep learning system for diabetic macular oedema classification on optical coherence tomography images"],"prefix":"10.1136","volume":"15","author":[{"ORCID":"https:\/\/orcid.org\/0009-0000-4394-240X","authenticated-orcid":false,"given":"Huishan","family":"Zhu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jie","family":"Ji","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jian-Wei","family":"Lin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ji","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yi","family":"Zheng","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Peiwen","family":"Xie","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Cui","family":"Liu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7863-7229","authenticated-orcid":false,"given":"Tsz Kin","family":"Ng","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jinqu","family":"Huang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yongqun","family":"Xiong","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hanfu","family":"Wu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Leixian","family":"Lin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9032-7274","authenticated-orcid":false,"given":"Mingzhi","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3298-7220","authenticated-orcid":false,"given":"Guihua","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"239","published-online":{"date-parts":[[2025,5,31]]},"reference":[{"key":"2025053106201112000_15.5.e099167.1","doi-asserted-by":"publisher","DOI":"10.1016\/S2213-8587(16)30052-3"},{"key":"2025053106201112000_15.5.e099167.2","unstructured":"International Diabetes Federation . IDF Diabetes Atlas, Available: https:\/\/diabetesatlas.org"},{"key":"2025053106201112000_15.5.e099167.3","doi-asserted-by":"publisher","DOI":"10.2337\/dc11-1909"},{"key":"2025053106201112000_15.5.e099167.4","doi-asserted-by":"publisher","DOI":"10.1016\/j.ophtha.2013.02.034"},{"key":"2025053106201112000_15.5.e099167.5","doi-asserted-by":"publisher","DOI":"10.1016\/0039-6257(84)90227-3"},{"key":"2025053106201112000_15.5.e099167.6","doi-asserted-by":"publisher","DOI":"10.1097\/IAE.0000000000000770"},{"key":"2025053106201112000_15.5.e099167.7","doi-asserted-by":"publisher","DOI":"10.1136\/bmj.311.7013.1131"},{"key":"2025053106201112000_15.5.e099167.8","doi-asserted-by":"publisher","DOI":"10.1136\/bjophthalmol-2013-304338"},{"key":"2025053106201112000_15.5.e099167.9","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pone.0014811","article-title":"SDOCT imaging to identify macular pathology in patients diagnosed with diabetic maculopathy by a digital photographic retinal screening programme","volume":"6","author":"Mackenzie","year":"2011","journal-title":"PLoS ONE"},{"key":"2025053106201112000_15.5.e099167.10","doi-asserted-by":"publisher","DOI":"10.12809\/hkmj166078"},{"key":"2025053106201112000_15.5.e099167.11","doi-asserted-by":"crossref","first-page":"222","DOI":"10.1001\/jamaophthalmol.2015.5332","article-title":"Comparison of prevalence of diabetic macular edema based on monocular fundus photography vs optical coherence tomography","volume":"134","author":"Wang","year":"2016","journal-title":"JAMA Ophthalmol"},{"key":"2025053106201112000_15.5.e099167.12","doi-asserted-by":"publisher","DOI":"10.1016\/j.ophtha.2003.06.028"},{"key":"2025053106201112000_15.5.e099167.13","article-title":"Optical coherence tomography (OCT) for detection of macular oedema in patients with diabetic retinopathy","volume":"2015","author":"Virgili","year":"2015","journal-title":"Cochrane Database Syst Rev"},{"key":"2025053106201112000_15.5.e099167.14","doi-asserted-by":"publisher","DOI":"10.1016\/j.ophtha.2018.04.007"},{"key":"2025053106201112000_15.5.e099167.15","doi-asserted-by":"crossref","DOI":"10.1186\/s40942-025-00638-9","article-title":"Automated and code-free development of a risk calculator using ChatGPT-4 for predicting diabetic retinopathy and macular edema without retinal imaging","volume":"11","author":"Choi","year":"2025","journal-title":"Int J Retina Vitreous"},{"key":"2025053106201112000_15.5.e099167.16","doi-asserted-by":"publisher","DOI":"10.1038\/nature14539"},{"key":"2025053106201112000_15.5.e099167.17","doi-asserted-by":"crossref","DOI":"10.3390\/diagnostics14010076","article-title":"Diabetic macular edema optical coherence tomography biomarkers detected with efficientNetV2B1 and ConvNeXt","volume":"14","author":"Suciu","year":"2023","journal-title":"Diagnostics (Basel)"},{"key":"2025053106201112000_15.5.e099167.18","doi-asserted-by":"publisher","DOI":"10.1097\/IAE.0000000000002992"},{"key":"2025053106201112000_15.5.e099167.19","doi-asserted-by":"crossref","unstructured":"Chan GCY , Kamble R , Muller H , et al . Fusing results of several deep learning architectures for automatic classification of normal and diabetic macular edema in optical coherence tomography. 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC); Honolulu, HI, 2018:670\u20133. doi:10.1109\/EMBC.2018.8512371","DOI":"10.1109\/EMBC.2018.8512371"},{"key":"2025053106201112000_15.5.e099167.20","doi-asserted-by":"crossref","DOI":"10.1155\/2016\/3298606","article-title":"Classification of SD-OCT Volumes Using Local Binary Patterns: Experimental Validation for DME Detection","volume":"2016","author":"Lema\u00eetre","year":"2016","journal-title":"J Ophthalmol"},{"key":"2025053106201112000_15.5.e099167.21","doi-asserted-by":"crossref","first-page":"109","DOI":"10.1016\/j.cmpb.2016.11.001","article-title":"An anomaly detection approach for the identification of DME patients using spectral domain optical coherence tomography images","volume":"139","author":"Sidib\u00e9","year":"2017","journal-title":"Comput Methods Programs Biomed"},{"key":"2025053106201112000_15.5.e099167.22","doi-asserted-by":"crossref","DOI":"10.1186\/s12938-017-0352-9","article-title":"Machine learning techniques for diabetic macular edema (DME) classification on SD-OCT images","volume":"16","author":"Alsaih","year":"2017","journal-title":"Biomed Eng Online"},{"key":"2025053106201112000_15.5.e099167.23","doi-asserted-by":"publisher","DOI":"10.2337\/dc20-3064"},{"key":"2025053106201112000_15.5.e099167.24","doi-asserted-by":"crossref","unstructured":"Hu Y , Ren J , Yang J , et al . Noise reduction by adaptive-SIN filtering for retinal OCT images. Sci Rep 2021;11. doi:10.1038\/s41598-021-98832-w","DOI":"10.1038\/s41598-021-98832-w"},{"key":"2025053106201112000_15.5.e099167.25","doi-asserted-by":"crossref","unstructured":"Zhou Z , Sodha V , Siddiquee MMR , et al . Models genesis: generic autodidactic models for 3D medical image analysis. Cham: Springer, 2019.","DOI":"10.1007\/978-3-030-32251-9_42"},{"key":"2025053106201112000_15.5.e099167.26","doi-asserted-by":"crossref","DOI":"10.1038\/s41467-019-13055-y","article-title":"Automated optimized parameters for T-distributed stochastic neighbor embedding improve visualization and analysis of large datasets","volume":"10","author":"Belkina","year":"2019","journal-title":"Nat Commun"},{"key":"2025053106201112000_15.5.e099167.27","doi-asserted-by":"crossref","unstructured":"Dmitry K , Philipp B . The art of using t-SNE for single-cell transcriptomics. Nat Commun 2019;10.","DOI":"10.1038\/s41467-019-13056-x"},{"key":"2025053106201112000_15.5.e099167.28","unstructured":"Springenberg JT , Dosovitskiy A , Brox T , et al . Striving for simplicity: the all convolutional net. 2014."},{"key":"2025053106201112000_15.5.e099167.29","doi-asserted-by":"publisher","DOI":"10.1001\/archopht.1996.01100140281004"},{"key":"2025053106201112000_15.5.e099167.30","doi-asserted-by":"publisher","DOI":"10.1016\/S0002-9394(99)00033-1"},{"key":"2025053106201112000_15.5.e099167.31","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); Venice. doi:10.1109\/ICCV.2017.74","DOI":"10.1109\/ICCV.2017.74"},{"key":"2025053106201112000_15.5.e099167.32","unstructured":"Bach S . Layer-wise relevance propagation for deep neural network architectures. Springer Singapore, 2016."},{"key":"2025053106201112000_15.5.e099167.33","doi-asserted-by":"crossref","unstructured":"Ribeiro MT , Singh S , Guestrin C . \u201cWhy should i trust you?\u201d: explaining the predictions of any classifier. Proceedings of the 2016 conference of the North American chapter of the association for computational linguistics; San Diego, California, 2016 doi:10.18653\/v1\/N16-3020 Available: http:\/\/aclweb.org\/anthology\/N16-3","DOI":"10.18653\/v1\/N16-3020"},{"key":"2025053106201112000_15.5.e099167.34","doi-asserted-by":"crossref","first-page":"51","DOI":"10.1016\/j.ajo.2016.09.002","article-title":"Real-world outcomes of ranibizumab treatment for diabetic macular edema in a United Kingdom national health service setting","volume":"172","author":"Patrao","year":"2016","journal-title":"Am J Ophthalmol"},{"key":"2025053106201112000_15.5.e099167.35","doi-asserted-by":"publisher","DOI":"10.1056\/NEJMoa1414264"}],"container-title":["BMJ Open"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1136\/bmjopen-2025-099167","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,5,31]],"date-time":"2025-05-31T13:20:25Z","timestamp":1748697625000},"score":1,"resource":{"primary":{"URL":"https:\/\/bmjopen.bmj.com\/lookup\/doi\/10.1136\/bmjopen-2025-099167"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,5]]},"references-count":35,"journal-issue":{"issue":"5","published-online":{"date-parts":[[2025,5,31]]},"published-print":{"date-parts":[[2025,5]]}},"alternative-id":["10.1136\/bmjopen-2025-099167"],"URL":"https:\/\/doi.org\/10.1136\/bmjopen-2025-099167","relation":{},"ISSN":["2044-6055","2044-6055"],"issn-type":[{"value":"2044-6055","type":"print"},{"value":"2044-6055","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,5]]}}}