{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,24]],"date-time":"2026-10-24T16:10:49Z","timestamp":1792858249120,"version":"4.5.0"},"reference-count":18,"publisher":"BMJ","funder":[{"DOI":"10.13039\/100015738","name":"Mary och Georg C. Ehrnrooths Stiftelse","doi-asserted-by":"publisher","award":["not available"],"award-info":[{"award-number":["not available"]}],"id":[{"id":"10.13039\/100015738","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["bmj.com"],"crossmark-restriction":true},"short-container-title":["Br J Ophthalmol"],"accepted":{"date-parts":[[2026,3,25]]},"abstract":"<jats:sec>\n                    <jats:title>Background\/aims<\/jats:title>\n                    <jats:p>Deep learning algorithms have shown promise for glaucoma detection using retinal imaging. The Northern Finland Birth Cohort Eye Study is a randomised prospective cohort designed to evaluate the efficacy and cost-effectiveness of glaucoma screening in a middle-aged population. This study assessed the performance and generalisability of G-RISK, a deep learning model for glaucoma screening from fundus images, in individuals aged 45\u201349 years with low glaucoma prevalence (1.1%).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>Four screening scenarios were evaluated: detection of (1) glaucoma eyes, (2) glaucoma patients, (3) glaucoma and glaucoma suspect eyes and (4) glaucoma and glaucoma suspect patients. Performance across four photographic modalities was assessed using the area under the receiver operating characteristic curve (AUC), the precision\u2013recall AUC (PR-AUC), and sensitivity at 95% specificity.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>G-RISK achieved its highest performance when screening for glaucoma eyes using colour fundus photographs, with an AUC of 0.83 (95% CI 0.75 to 0.90), PR-AUC of 0.16 (95% CI 0.05 to 0.28) and sensitivity at 95% specificity of 0.43 (95% CI 0.28 to 0.59). Furthermore, discrimination remained consistent across other modalities (AUC 0.75\u20130.78). Performance was higher for confirmed glaucoma than combined glaucoma and glaucoma suspect cases (AUC 0.83 vs 0.67).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>G-RISK achieved the best results on colour fundus images, while maintaining discrimination across other imaging modalities. Screening metrics were constrained by the relatively young age and low glaucoma prevalence of the cohort. The reduced performance observed when including glaucoma suspects may relate to variability and uncertainty in their clinical definition.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1136\/bjo-2025-328523","type":"journal-article","created":{"date-parts":[[2026,4,15]],"date-time":"2026-04-15T17:20:21Z","timestamp":1776273621000},"page":"bjo-2025-328523","update-policy":"https:\/\/doi.org\/10.1136\/crossmarkpolicy","source":"Crossref","is-referenced-by-count":1,"title":["Deep learning in population-based screening for glaucoma"],"prefix":"10.1136","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-1182-7954","authenticated-orcid":false,"given":"Joel","family":"Pitk\u00e4nen","sequence":"first","affiliation":[{"name":"Department of Ophthalmology and Medical Research Center, Oulu University Hospital, Oulu, Finland"},{"name":"Research Unit of Clinical Medicine, University of Oulu, Oulu, Finland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3214-4247","authenticated-orcid":false,"given":"Ruben","family":"Hemelings","sequence":"additional","affiliation":[{"name":"Singapore Eye Research Institute, Singapore National Eye Centre, Singapore"},{"name":"Research Group Ophthalmology, Department of Neurosciences, KU Leuven, Leuven, Belgium"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Oona","family":"Ahokas","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology and Medical Research Center, Oulu University Hospital, Oulu, Finland"},{"name":"Department of Mathematical Sciences, University of Oulu, Oulu, Finland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3424-7207","authenticated-orcid":false,"given":"Johanna M","family":"Liinamaa","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology and Medical Research Center, Oulu University Hospital, Oulu, Finland"},{"name":"Research Unit of Clinical Medicine, University of Oulu, Oulu, Finland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6521-6695","authenticated-orcid":false,"given":"Elina","family":"Karvonen","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology, Helsinki University Hospital, Helsinki, Uusimaa, Finland"},{"name":"Department of Ophthalmology, University of Helsinki, Helsinki, Uusimaa, Finland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Katri","family":"Stoor","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology and Medical Research Center, Oulu University Hospital, Oulu, Finland"},{"name":"Research Unit of Clinical Medicine, University of Oulu, Oulu, Finland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Marja","family":"Honkamo","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology and Medical Research Center, Oulu University Hospital, Oulu, Finland"},{"name":"Research Unit of Clinical Medicine, University of Oulu, Oulu, Finland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Pasi","family":"H\u00e4gg","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology and Medical Research Center, Oulu University Hospital, Oulu, Finland"},{"name":"Research Unit of Clinical Medicine, University of Oulu, Oulu, Finland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Anja","family":"Tuulonen","sequence":"additional","affiliation":[{"name":"Tays Eye Center, Tampere University Hospital, Tampere, Pirkanmaa, Finland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ingeborg","family":"Stalmans","sequence":"additional","affiliation":[{"name":"Research Group Ophthalmology, Department of Neurosciences, KU Leuven, Leuven, Belgium"},{"name":"Department of Ophthalmology, University Hospitals UZ Leuven, Leuven, Belgium"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ville","family":"Saarela","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology and Medical Research Center, Oulu University Hospital, Oulu, Finland"},{"name":"Research Unit of Clinical Medicine, University of Oulu, Oulu, Finland"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"239","published-online":{"date-parts":[[2026,4,15]]},"reference":[{"key":"2026041510201627000_bjo-2025-328523v1.1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ophtha.2014.05.013"},{"key":"2026041510201627000_bjo-2025-328523v1.2","first-page":"741","article-title":"Number of ganglion cells in glaucoma eyes compared with threshold visual field tests in the same persons","volume":"41","author":"Kerrigan-Baumrind","year":"2000","journal-title":"Invest Ophthalmol Vis Sci"},{"key":"2026041510201627000_bjo-2025-328523v1.3","doi-asserted-by":"publisher","DOI":"10.1136\/bjophthalmol-2019-314795"},{"key":"2026041510201627000_bjo-2025-328523v1.4","doi-asserted-by":"publisher","DOI":"10.1001\/archopht.118.9.1187"},{"key":"2026041510201627000_bjo-2025-328523v1.5","doi-asserted-by":"crossref","first-page":"206","DOI":"10.1016\/j.media.2017.04.007","article-title":"A machine-learning graph-based approach for 3D segmentation of Bruch\u2019s membrane opening from glaucomatous SD-OCT volumes","volume":"39","author":"Miri","year":"2017","journal-title":"Med Image Anal"},{"key":"2026041510201627000_bjo-2025-328523v1.6","doi-asserted-by":"publisher","DOI":"10.1016\/j.ophtha.2018.12.033"},{"key":"2026041510201627000_bjo-2025-328523v1.7","doi-asserted-by":"publisher","DOI":"10.1016\/j.ajo.2019.01.011"},{"key":"2026041510201627000_bjo-2025-328523v1.8","doi-asserted-by":"publisher","DOI":"10.1111\/aos.14193"},{"key":"2026041510201627000_bjo-2025-328523v1.9","doi-asserted-by":"crossref","DOI":"10.1038\/s41598-021-99605-1","article-title":"Deep learning on fundus images detects glaucoma beyond the optic disc","volume":"11","author":"Hemelings","year":"2021","journal-title":"Sci Rep"},{"key":"2026041510201627000_bjo-2025-328523v1.10","doi-asserted-by":"crossref","DOI":"10.1038\/s41746-023-00857-0","article-title":"A generalizable deep learning regression model for automated glaucoma screening from fundus images","volume":"6","author":"Hemelings","year":"2023","journal-title":"NPJ Digit Med"},{"key":"2026041510201627000_bjo-2025-328523v1.11","doi-asserted-by":"crossref","first-page":"200","DOI":"10.1111\/aos.13912","article-title":"Prevalence of glaucoma in the Northern Finland Birth Cohort Eye Study","volume":"97","author":"Karvonen","year":"2019","journal-title":"Acta Ophthalmol"},{"key":"2026041510201627000_bjo-2025-328523v1.12","doi-asserted-by":"publisher","DOI":"10.1034\/j.1600-0420.2003.00021.x"},{"key":"2026041510201627000_bjo-2025-328523v1.13","doi-asserted-by":"crossref","DOI":"10.1186\/1471-2415-13-51","article-title":"The Northern Finland Birth Cohort Eye Study: Design and baseline characteristics","volume":"13","author":"Saarela","year":"2013","journal-title":"BMC Ophthalmol"},{"key":"2026041510201627000_bjo-2025-328523v1.14","doi-asserted-by":"publisher","DOI":"10.2307\/2531595"},{"key":"2026041510201627000_bjo-2025-328523v1.15","doi-asserted-by":"crossref","first-page":"1695","DOI":"10.1007\/s00417-018-4011-z","article-title":"Prevalence of glaucoma in Germany: results from the Gutenberg Health Study","volume":"256","author":"H\u00f6hn","year":"2018","journal-title":"Graefes Arch Clin Exp Ophthalmol"},{"key":"2026041510201627000_bjo-2025-328523v1.16","doi-asserted-by":"publisher","DOI":"10.1016\/S0161-6420(96)30449-1"},{"key":"2026041510201627000_bjo-2025-328523v1.17","doi-asserted-by":"crossref","first-page":"100916","DOI":"10.1016\/j.landig.2025.100916","article-title":"Evaluation of performance measures in predictive artificial intelligence models to support medical decisions: overview and guidance","volume":"7","author":"Van Calster","year":"2025","journal-title":"Lancet Digit Health"},{"key":"2026041510201627000_bjo-2025-328523v1.18","doi-asserted-by":"crossref","unstructured":"Angelotti G , Gallifant J , Hansen L , et al . A closer look at AUROC and AUPRC under class imbalance. In: Globerson A , Mackey L , Belgrave D , et al , eds. Advances in neural information processing systems. Curran Associates, Inc, 2024: 44102\u201363.","DOI":"10.52202\/079017-1400"}],"container-title":["British Journal of Ophthalmology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1136\/bjo-2025-328523","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,15]],"date-time":"2026-04-15T17:20:23Z","timestamp":1776273623000},"score":1,"resource":{"primary":{"URL":"https:\/\/bjo.bmj.com\/lookup\/doi\/10.1136\/bjo-2025-328523"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,4,15]]},"references-count":18,"alternative-id":["10.1136\/bjo-2025-328523"],"URL":"https:\/\/doi.org\/10.1136\/bjo-2025-328523","relation":{},"ISSN":["0007-1161","1468-2079"],"issn-type":[{"value":"0007-1161","type":"print"},{"value":"1468-2079","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,4,15]]}}}