{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,30]],"date-time":"2026-09-30T22:55:58Z","timestamp":1790808958822,"version":"4.1.0"},"reference-count":26,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,11,27]],"date-time":"2025-11-27T00:00:00Z","timestamp":1764201600000},"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>Cataracts, the leading cause of reversible blindness globally, require timely detection and intervention for effective prevention of blindness. Artificial intelligence can assist in massive screening, however, existing models often trained on homogeneous, single-center data, suer from poor generalizability.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>To address this challenge, we developed and validated a deep learning model trained on a large-scale, multicenter, real-world dataset comprising 22,094 slit-lamp images from 21 ophthalmic institutions across 12 provinces and municipalities in China. We designed a cascaded framework that emulates the sequential reasoning of a clinical diagnostic workflow, a methodological approach for ensuring reliability on noisy, real-world data. It first performs an automated quality assessment, then screens for common confounders like pterygium, and finally conducts a differential diagnosis among cataract, post-cataract surgery, other ocular diseases, and healthy eyes. Within this framework, we evaluated several deep learning architectures.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>In the cataract classification task, the leading models demonstrated excellent performance on an independent test set. For instance, the ResNet50-IBN based model achieved an accuracy of 93.74%, specificity of 97.74% and an area under the curve (AUC) of 95.30%.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>This study demonstrates that training on multicenter, real-world data yields a robust and generalizable model, providing a powerful tool for largescale ophthalmic screening. Specifically, our model establishes a methodological blueprint for developing trustworthy medical deep learning systems.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fmed.2025.1691419","type":"journal-article","created":{"date-parts":[[2025,11,27]],"date-time":"2025-11-27T06:30:27Z","timestamp":1764225027000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":2,"title":["A deep learning-driven cataract screening model derived from multicenter real-world dataset"],"prefix":"10.3389","volume":"12","author":[{"given":"Zhonghui","family":"Cui","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yu","family":"Cheng","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Siqi","family":"Pan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yong","family":"Zhu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Weiwei","family":"Dai","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2025,11,27]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"S22","DOI":"10.1016\/S0140-6736(19)32358-X","article-title":"A universal artificial intelligence platform for collaborative management of cataracts","volume":"394","author":"Wu","year":"2019","journal-title":"Lancet"},{"key":"B2","doi-asserted-by":"publisher","first-page":"e144","DOI":"10.1016\/S2214-109X(20)30489-7","article-title":"Causes of blindness and vision impairment in 2020 and trends over 30 years, and prevalence of avoidable blindness in relation to VISION 2020: the Right to Sight: an analysis for the Global Burden of Disease Study","volume":"9","author":"Steinmetz","year":"2021","journal-title":"Lancet Glob Health"},{"key":"B3","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/nrdp.2015.14","article-title":"Cataract","volume":"1","author":"Lam","year":"2015","journal-title":"Nat Rev Dis Primers"},{"key":"B4","doi-asserted-by":"publisher","first-page":"2169","DOI":"10.1016\/j.ophtha.2015.06.052","article-title":"Ancestry, socioeconomic status, and age-related cataract in Asians: the Singapore epidemiology of eye diseases study","volume":"122","author":"Chua","year":"2015","journal-title":"Ophthalmology"},{"key":"B5","doi-asserted-by":"publisher","first-page":"991","DOI":"10.1001\/archopht.1989.01070020053028","article-title":"Lens opacities classification system II (LOCS II)","volume":"107","author":"Chylack","year":"1989","journal-title":"Arch Ophthalmol"},{"key":"B6","doi-asserted-by":"publisher","first-page":"1428","DOI":"10.1016\/S0161-6420(90)32391-6","article-title":"Assessment of cataracts from photographs in the Beaver Dam Eye Study","volume":"97","author":"Klein","year":"1990","journal-title":"Ophthalmology"},{"key":"B7","doi-asserted-by":"publisher","first-page":"86","DOI":"10.1016\/j.optom.2014.07.006","article-title":"Mart\u00ednez CP. Refractive changes in nuclear, cortical and posterior subcapsular cataracts effect of the type and grade","volume":"8","author":"Ajenjo","year":"2015","journal-title":"J Optom"},{"key":"B8","doi-asserted-by":"publisher","first-page":"81","DOI":"10.1111\/j.1442-9071.1989.tb00492.x","article-title":"The clinical grading of lens opacities","volume":"17","author":"Taylor","year":"1989","journal-title":"Aust N Z J Ophthalmol"},{"key":"B9","doi-asserted-by":"publisher","first-page":"e489","DOI":"10.25259\/IHOPEJO_15_2021","article-title":"The Lancet Global Health commission on global eye health: vision beyond 2020","volume":"9","author":"Burton","year":"2021","journal-title":"Lancet Glob Health"},{"key":"B10","doi-asserted-by":"crossref","first-page":"83","DOI":"10.1109\/CAI59869.2024.00024","article-title":"Accurate and explainable cataract detection using eye images taken by hand-held slit-lamp cameras","volume-title":"2024 IEEE Conference on Artificial Intelligence (CAI)","author":"Fung","year":"2024"},{"key":"B11","doi-asserted-by":"publisher","first-page":"100147","DOI":"10.1016\/j.xops.2022.100147","article-title":"Deep learning-based cataract detection and grading from slit-lamp and retro-illumination photographs: model development and validation study","volume":"2","author":"Son","year":"2022","journal-title":"Ophthalmol Sci"},{"key":"B12","doi-asserted-by":"publisher","first-page":"290","DOI":"10.1145\/3632410.3632440","article-title":"Multimodal and multitask approaches for cataract grading","author":"Saxena","year":"2024","journal-title":"Proceedings of the 7th Joint International Conference on Data Science & Management of Data (11th ACM IKDD CODS and 29th COMAD)"},{"key":"B13","doi-asserted-by":"publisher","first-page":"174169","DOI":"10.1109\/ACCESS.2020.3025346","article-title":"Unified diagnosis framework for automated nuclear cataract grading based on smartphone slit-lamp images","volume":"8","author":"Hu","year":"2020","journal-title":"IEEE Access"},{"key":"B14","article-title":"Aiming to Scale: The Validity, Usability, and Business Model of A Novel Teleophthalmic Deep Learning Screening Technology","author":"He","year":"2020"},{"key":"B15","doi-asserted-by":"publisher","first-page":"88","DOI":"10.1097\/01.APO.0000656988.16221.04","article-title":"Artificial intelligence for cataract detection and management","volume":"9","author":"Goh","year":"2020","journal-title":"Asia-Pac J Ophthalmol"},{"key":"B16","doi-asserted-by":"publisher","first-page":"464","DOI":"10.1007\/978-3-030-01225-0_29","article-title":"Two at once: enhancing learning and generalization capacities via IBN-Net","author":"Pan","year":"2018","journal-title":"Proceedings of the European Conference on Computer Vision (ECCV)"},{"key":"B17","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/s12916-019-1426-2","article-title":"Key challenges for delivering clinical impact with artificial intelligence","volume":"17","author":"Kelly","year":"2019","journal-title":"BMC Med"},{"key":"B18","doi-asserted-by":"publisher","first-page":"325","DOI":"10.1007\/978-3-030-01261-8_20","article-title":"Bisenet: bilateral segmentation network for real-time semantic segmentation","author":"Yu","year":"2018","journal-title":"Proceedings of the European Conference on Computer Vision (ECCV)"},{"key":"B19","doi-asserted-by":"publisher","first-page":"4700","DOI":"10.1109\/CVPR.2017.243","article-title":"Densely connected convolutional networks","author":"Huang","year":"2017","journal-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition"},{"key":"B20","first-page":"6105","article-title":"Efficientnet: rethinking model scaling for convolutional neural networks","volume-title":"International Conference on Machine Learning","author":"Tan","year":"2019"},{"key":"B21","first-page":"11976","article-title":"A convnet for the 2020s","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Liu","year":"2022"},{"key":"B22","doi-asserted-by":"publisher","first-page":"e54047","DOI":"10.2196\/54047","article-title":"Slit lamp report generation and question answering: development and validation of a multimodal transformer model with large language model integration","volume":"26","author":"Zhao","year":"2024","journal-title":"J Med Internet Res"},{"key":"B23","doi-asserted-by":"publisher","first-page":"1089","DOI":"10.1109\/TMI.2023.3327274","article-title":"A transformer-based knowledge distillation network for cortical cataract grading","volume":"43","author":"Wang","year":"2023","journal-title":"IEEE Trans Med Imaging"},{"key":"B24","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1016\/j.patter.2024.100928","article-title":"FedEYE: A scalable and flexible end-to-end federated learning platform for ophthalmology","author":"Yan","year":"2024"},{"key":"B25","doi-asserted-by":"publisher","first-page":"23","DOI":"10.1167\/iovs.62.7.23","article-title":"Early onset of age-related cataracts in cystine\/glutamate antiporter knockout mice","volume":"62","author":"Martis RM Li","year":"2021","journal-title":"Invest Ophthalmol Vis Sci"},{"key":"B26","doi-asserted-by":"publisher","first-page":"6","DOI":"10.3390\/jfb14010006","article-title":"Reduction in lens epithelial cell senescence burden through Dasatinib plus quercetin or rapamycin alleviates D-galactose-induced cataract progression","volume":"14","author":"Wang","year":"2022","journal-title":"J Funct Biomater"}],"container-title":["Frontiers in Medicine"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2025.1691419\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,27]],"date-time":"2025-11-27T06:30:29Z","timestamp":1764225029000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2025.1691419\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,27]]},"references-count":26,"alternative-id":["10.3389\/fmed.2025.1691419"],"URL":"https:\/\/doi.org\/10.3389\/fmed.2025.1691419","relation":{},"ISSN":["2296-858X"],"issn-type":[{"value":"2296-858X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,11,27]]},"article-number":"1691419"}}