{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,24]],"date-time":"2026-09-24T07:00:38Z","timestamp":1790233238209,"version":"4.1.0"},"reference-count":31,"publisher":"IOP Publishing","issue":"9","license":[{"start":{"date-parts":[[2026,9,24]],"date-time":"2026-09-24T00:00:00Z","timestamp":1790208000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/publishingsupport.iopscience.iop.org\/iop-standard\/v1"},{"start":{"date-parts":[[2026,9,24]],"date-time":"2026-09-24T00:00:00Z","timestamp":1790208000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"funder":[{"name":"National Science Foundation","award":["ITEST 2148451"],"award-info":[{"award-number":["ITEST 2148451"]}]},{"name":"National Institute of Health","award":["Bridge2AI OT2OD032701"],"award-info":[{"award-number":["Bridge2AI OT2OD032701"]}]}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Physiol. Meas."],"published-print":{"date-parts":[[2026,9,30]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>\n                    <jats:italic>Objective.<\/jats:italic>\n                    Diabetic retinopathy (DR) is the leading cause of preventable blindness in adults and poses significant challenges in low- and middle-income regions due to limited access to skilled clinicians and diagnostic facilities. Automated screening solutions using artificial intelligence (AI) have emerged as an efficient alternative, achieving high diagnostic accuracy. However, these solutions are often developed using data from specific populations obtained using relatively expensive high-end devices. This study addresses the potential scope limitations by evaluating the AI-based screening tool\n                    <jats:italic>retina.help<\/jats:italic>\n                    across eight datasets representing diverse populations, imaging modalities, and geographic regions.\n                    <jats:italic>Approach.<\/jats:italic>\n                    The datasets include both public and private sources, with images captured using tabletop and handheld fundus cameras. Key performance metrics for detecting binary referable DR\u2014sensitivity, specificity, and area under the receiver operating characteristic curve (AUROC)\u2014were calculated on an image-by-image basis.\n                    <jats:italic>Main results.<\/jats:italic>\n                    <jats:italic>Retina.help<\/jats:italic>\n                    demonstrated high accuracy on tabletop images, achieving AUROC values of 0.97 on the BRSET and DeepDRiD datasets. Handheld device performance was more variable, with AUROC ranging from 0.88 (Filipino dataset) to 0.99 (Finnish dataset). Sensitivity declined with increased retinal pigmentation, as evidenced by lower values for datasets from Tanzania (62.2%) and Brazil (76.7%) compared to Finland (89.9%). Images from handheld devices often yielded lower sensitivity due to challenges related to low-contrast images. Nonetheless,\n                    <jats:italic>retina.help<\/jats:italic>\n                    generalized well across diverse datasets, showcasing its robustness.\n                    <jats:italic>Significance.<\/jats:italic>\n                    The study highlights the impact of imaging equipment, demographics, and image quality on diagnostic performance. These findings underscore the need for benchmarking AI-based DR screening tools using standardized datasets that encompass diverse populations and imaging conditions. Such evaluations can guide the development of equitable, reliable and robust screening solutions.\n                  <\/jats:p>","DOI":"10.1088\/1361-6579\/aea2ee","type":"journal-article","created":{"date-parts":[[2026,9,4]],"date-time":"2026-09-04T22:53:20Z","timestamp":1788562400000},"page":"095007","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Diabetic retinopathy screening using artificial intelligence: a comprehensive performance comparison on datasets from diverse populations and imaging modalities"],"prefix":"10.1088","volume":"47","author":[{"ORCID":"https:\/\/orcid.org\/0009-0008-0943-6499","authenticated-orcid":true,"given":"Philipp","family":"Schapotschnikow","sequence":"first","affiliation":[{"name":"Brains for hire UG"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-5122-8005","authenticated-orcid":true,"given":"Vladyslav","family":"Dusiak","sequence":"additional","affiliation":[{"name":"Brains for hire UG"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"data-curation","vocabulary":"credit"},{"role":"methodology","vocabulary":"credit"},{"role":"software","vocabulary":"credit"}]},{"given":"Murat","family":"Osswald","sequence":"additional","affiliation":[{"name":"Brains for hire UG"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"data-curation","vocabulary":"credit"},{"role":"methodology","vocabulary":"credit"},{"role":"software","vocabulary":"credit"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8024-9348","authenticated-orcid":true,"given":"Paolo S","family":"Silva","sequence":"additional","affiliation":[{"name":"Joslin Diabetes Center Beetham Eye Institute"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"conceptualization","vocabulary":"credit"},{"role":"resources","vocabulary":"credit"},{"role":"supervision","vocabulary":"credit"}]},{"given":"Leo Anthony","family":"Celi","sequence":"additional","affiliation":[{"name":"Massachusetts Institute of Technology"},{"name":"Harvard T.H. Chan School of Public Health"},{"name":"Beth Israel Deaconess Medical Center"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"conceptualization","vocabulary":"credit"},{"role":"resources","vocabulary":"credit"},{"role":"supervision","vocabulary":"credit"}]},{"given":"Luis Filipe","family":"Nakayama","sequence":"additional","affiliation":[{"name":"Massachusetts Institute of Technology"},{"name":"Sao Paulo Federal University"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"conceptualization","vocabulary":"credit"},{"role":"resources","vocabulary":"credit"},{"role":"supervision","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"}]}],"member":"266","published-online":{"date-parts":[[2026,9,24]]},"reference":[{"key":"pmeaaea2eebib1","doi-asserted-by":"publisher","first-page":"351","DOI":"10.1001\/jamaophthalmol.2013.1743","type":"journal-article","article-title":"Automated analysis of retinal images for detection of referable diabetic retinopathy","volume":"131","author":"Abr\u00e0moff","year":"2013","journal-title":"JAMA Ophthalmol."},{"key":"pmeaaea2eebib2","article-title":"Ocular disease recognition (ODIR)","author":"Andrew","year":"n.d.","type":"web-resource"},{"key":"pmeaaea2eebib3","doi-asserted-by":"publisher","first-page":"13","DOI":"10.1167\/tvst.10.2.13","type":"journal-article","article-title":"Addressing artificial intelligence bias in retinal disease diagnostics","volume":"10","author":"Burlina","year":"2021","journal-title":"Transl. Vis. Sci. Technol."},{"key":"pmeaaea2eebib4","doi-asserted-by":"publisher","author":"Decenci\u00e8re","year":"2013","DOI":"10.1016\/j.irbm.2013.01.010)","type":"book"},{"key":"pmeaaea2eebib5","article-title":"Diabetic retinopathy detection","author":"Dugas","year":"2015","type":"web-resource"},{"key":"pmeaaea2eebib6","doi-asserted-by":"publisher","first-page":"2352","DOI":"10.3390\/jcm10112352","type":"journal-article","article-title":"Analysis and comparison of two artificial intelligence diabetic retinopathy screening algorithms in a pilot study: IDx-DR and retinalyze","volume":"10","author":"Grzybowski","year":"2021","journal-title":"J. Clin. Med."},{"key":"pmeaaea2eebib7","doi-asserted-by":"publisher","first-page":"633","DOI":"10.1007\/s40123-023-00652-w","type":"journal-article","article-title":"Approval and certification of ophthalmic AI devices in the European Union","volume":"12","author":"Grzybowski","year":"2023","journal-title":"Ophthalmol. Ther."},{"key":"pmeaaea2eebib8","doi-asserted-by":"publisher","first-page":"451","DOI":"10.1038\/s41433-019-0566-0","type":"journal-article","article-title":"Artificial intelligence for diabetic retinopathy screening: a review","volume":"34","author":"Grzybowski","year":"2020","journal-title":"Eye"},{"key":"pmeaaea2eebib9","author":"Grzybowski","year":"2024","type":"book"},{"key":"pmeaaea2eebib10","doi-asserted-by":"publisher","first-page":"703","DOI":"10.1016\/j.oret.2023.03.003","type":"journal-article","article-title":"Performance of automated machine learning for diabetic retinopathy image classification from multi-field handheld retinal images","volume":"7","author":"Jacoba","year":"2023a","journal-title":"Ophthalmol. Retina"},{"key":"pmeaaea2eebib11","first-page":"255","type":"journal-article","article-title":"Ethnic diversity of retinal images used to train artificial intelligence models improves diagnostic accuracy to detect diabetic retinopathy","volume":"64","author":"Jacoba","year":"2023b","journal-title":"Invest. Ophthalmol. Vis. Sci."},{"key":"pmeaaea2eebib12","article-title":"APTOS 2019 blindness detection","author":"Karthik","year":"2019","type":"web-resource"},{"key":"pmeaaea2eebib13","doi-asserted-by":"publisher","first-page":"3587","DOI":"10.3390\/jcm12103587","type":"journal-article","article-title":"A global review of publicly available datasets containing fundus images","volume":"12","author":"Krzywicki","year":"2023","journal-title":"J. Clin. Med."},{"key":"pmeaaea2eebib14","doi-asserted-by":"publisher","DOI":"10.1080\/07853890.2024.2352018","type":"journal-article","article-title":"Comparison of 21 artificial intelligence algorithms in automated diabetic retinopathy screening using handheld fundus camera","volume":"56","author":"Kubin","year":"2024","journal-title":"Ann. Med."},{"key":"pmeaaea2eebib15","doi-asserted-by":"publisher","first-page":"1168","DOI":"10.2337\/dc20-1877","type":"journal-article","article-title":"Multicenter, head-to-head, real-world validation study of seven automated artificial intelligence diabetic retinopathy screening systems","volume":"44","author":"Lee","year":"2021","journal-title":"Diabetes Care"},{"key":"pmeaaea2eebib16","doi-asserted-by":"publisher","first-page":"511","DOI":"10.1016\/j.ins.2019.06.011","type":"journal-article","article-title":"Diagnostic assessment of deep learning algorithms for diabetic retinopathy screening","volume":"501","author":"Li","year":"2019","journal-title":"Inf. Sci."},{"key":"pmeaaea2eebib17","doi-asserted-by":"publisher","DOI":"10.1016\/j.patter.2022.100512","type":"journal-article","article-title":"DeepDRiD: diabetic retinopathy grading and image quality estimation challenge","volume":"3","author":"Liu","year":"2022","journal-title":"Patterns"},{"key":"pmeaaea2eebib18","doi-asserted-by":"publisher","first-page":"747","DOI":"10.1001\/jamaophthalmol.2023.2289","type":"journal-article","article-title":"Prevalence of diabetic retinopathy in the US in 2021","volume":"141","author":"Lundeen","year":"2023","journal-title":"JAMA Ophthalmol."},{"key":"pmeaaea2eebib19","author":"Malerbi","year":"2024","type":"book"},{"key":"pmeaaea2eebib20","article-title":"DRStageNet: deep learning for diabetic retinopathy staging from fundus images","author":"Men","year":"2024","type":"preprint"},{"key":"pmeaaea2eebib21","doi-asserted-by":"publisher","first-page":"1410","DOI":"10.1177\/19322968211029943","type":"journal-article","article-title":"The challenge of diabetic retinopathy standardization in an ophthalmological dataset","volume":"15","author":"Nakayama","year":"2021","journal-title":"J. Diabetes Sci. Technol."},{"key":"pmeaaea2eebib22","doi-asserted-by":"publisher","first-page":"2","DOI":"10.13026\/xcxw-8198","type":"journal-article","article-title":"A Brazilian multilabel ophthalmological dataset (BRSET)","volume":"13026","author":"Nakayama","year":"2023","journal-title":"PhysioNet"},{"key":"pmeaaea2eebib23","doi-asserted-by":"publisher","DOI":"10.13026\/qxpd-1y65","type":"journal-article","article-title":"mBRSET, a mobile Brazilian retinal dataset (version 1.0)","author":"Nakayama","year":"2024","journal-title":"PhysioNet"},{"key":"pmeaaea2eebib24","doi-asserted-by":"publisher","first-page":"203","DOI":"10.1016\/S0140-6736(23)02044-5","type":"journal-article","article-title":"Global, regional, and national burden of diabetes from 1990 to 2021","volume":"402","author":"Ong","year":"2021","journal-title":"Lancet"},{"key":"pmeaaea2eebib25","author":"Porwal","year":"2018","type":"book"},{"key":"pmeaaea2eebib26","doi-asserted-by":"publisher","DOI":"10.1016\/j.xops.2023.100457","type":"journal-article","article-title":"Accuracy of integrated artificial intelligence grading using handheld retinal imaging","volume":"4","author":"Salongcay","year":"2023","journal-title":"Ophthalmol. Sci."},{"key":"pmeaaea2eebib27","first-page":"6105","type":"conference-proceedings","article-title":"EfficientNet: rethinking model scaling for convolutional neural networks","volume":"vol 97","author":"Tan","year":"2019"},{"key":"pmeaaea2eebib28","first-page":"7403","type":"conference-proceedings","article-title":"Learn to segment retinal lesions and beyond","author":"Wei","year":"2020"},{"key":"pmeaaea2eebib29","doi-asserted-by":"publisher","first-page":"1677","DOI":"10.1016\/S0161-6420(03)00475-5","type":"journal-article","article-title":"Proposed international clinical diabetic retinopathy severity scales","volume":"110","author":"Wilkinson","year":"2003","journal-title":"Ophthalmology"},{"key":"pmeaaea2eebib30","doi-asserted-by":"publisher","first-page":"1608","DOI":"10.1016\/j.ophtha.2018.04.007","type":"journal-article","article-title":"Guidelines on diabetic eye care","volume":"125","author":"Wong","year":"2018","journal-title":"Ophthalmology"},{"key":"pmeaaea2eebib31","doi-asserted-by":"publisher","first-page":"556","DOI":"10.2337\/dc11-1909","type":"journal-article","article-title":"Global prevalence and major risk factors of diabetic retinopathy","volume":"35","author":"Yau","year":"2012","journal-title":"Diabetes Care"}],"container-title":["Physiological Measurement"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6579\/aea2ee","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6579\/aea2ee\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6579\/aea2ee","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6579\/aea2ee\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6579\/aea2ee\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6579\/aea2ee\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6579\/aea2ee\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6579\/aea2ee\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,24]],"date-time":"2026-09-24T06:18:54Z","timestamp":1790230734000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6579\/aea2ee"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,24]]},"references-count":31,"journal-issue":{"issue":"9","published-online":{"date-parts":[[2026,9,24]]},"published-print":{"date-parts":[[2026,9,30]]}},"URL":"https:\/\/doi.org\/10.1088\/1361-6579\/aea2ee","relation":{},"ISSN":["0967-3334","1361-6579"],"issn-type":[{"value":"0967-3334","type":"print"},{"value":"1361-6579","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,24]]},"assertion":[{"value":"Diabetic retinopathy screening using artificial intelligence: a comprehensive performance comparison on datasets from diverse populations and imaging modalities","name":"article_title","label":"Article Title"},{"value":"Physiological Measurement","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2026 Institute of Physics and Engineering in Medicine. All rights, including for text and data mining, AI training, and similar technologies, are reserved.","name":"copyright_information","label":"Copyright Information"},{"value":"2026-01-12","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2026-09-04","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2026-09-24","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}