{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,5,29]],"date-time":"2026-05-29T21:10:56Z","timestamp":1780089056102,"version":"3.54.0"},"posted":{"date-parts":[[2026,5,27]]},"group-title":"Ophthalmology","reference-count":29,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,5,27]],"date-time":"2026-05-27T00:00:00Z","timestamp":1779840000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nd\/4.0\/"}],"funder":[{"name":"NIH","award":["R01 EY036222"],"award-info":[{"award-number":["R01 EY036222"]}]},{"name":"NIH","award":["R21 EY035298"],"award-info":[{"award-number":["R21 EY035298"]}]},{"name":"NIH","award":["P30 EY003790"],"award-info":[{"award-number":["P30 EY003790"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,5,27]]},"abstract":"<jats:title>ABSTRACT<\/jats:title>\n                <jats:sec>\n                  <jats:title>Purpose<\/jats:title>\n                  <jats:p>To predict retinal nerve fiber layer thickness (RNFLT) norms from fundus images.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Methods<\/jats:title>\n                  <jats:p>\n                    We selected 18,000 OCT scans and visual fields (VF) from the Massachusetts Eye and Ear Glaucoma Service. A U-Net-based deep learning model was developed to predict RNFLT norms from OCT en face fundus images. A total of 10,000 OCT scans with normal VFs (mean deviation [MD] \u2265 -1 dB, glaucoma hemifield test within normal limits, and pattern standard deviation probability &gt; 5%) tested within 30 days were used for training, while the remaining 8,000 OCT scans (mean VF MD: \u22123.3 \u00b1 4.9 dB), including 2,419 scans with normal VFs, were used for evaluation. Structure-function correlations between RNFLT maps and VFs were assessed using linear regression and VGG-16 across original RNFLT maps, deviation maps, and their combination. Performance was evaluated using correlation coefficients, mean absolute error (MAE), and R\n                    <jats:sup>2<\/jats:sup>\n                    .\n                  <\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>\n                    Predicted RNFLT norm maps showed agreement with baseline RNFLT maps in eyes with normal VFs (R\n                    <jats:sup>2<\/jats:sup>\n                    = 0.81 \u00b1 0.13). RNFLT deviation maps correlated more strongly with VF MD than original RNFLT maps (R = 0.42 vs. 0.19, p &lt; 0.01). In deep learning-based VF prediction, combining original and deviation maps achieved the best performance (MAE = 3.31 dB, R\n                    <jats:sup>2<\/jats:sup>\n                    = 0.39), outperforming the model (p &lt; 0.05) using original RNFLT maps alone (MAE = 3.36 dB, R\n                    <jats:sup>2<\/jats:sup>\n                    = 0.35).\n                  <\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Conclusions<\/jats:title>\n                  <jats:p>Deep learning can estimate individualized RNFLT norms and improve structure-function assessment in glaucoma.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Translational Relevance<\/jats:title>\n                  <jats:p>Personalized RNFLT norm prediction may improve detection of glaucomatous damage.<\/jats:p>\n                <\/jats:sec>","DOI":"10.64898\/2026.05.26.26354081","type":"posted-content","created":{"date-parts":[[2026,5,27]],"date-time":"2026-05-27T17:55:15Z","timestamp":1779904515000},"source":"Crossref","is-referenced-by-count":0,"title":["Deep Learning Prediction of Personalized Peripapillary Retinal Nerve Fiber Layer Thickness Norms from Fundus Images in Glaucoma"],"prefix":"10.64898","author":[{"given":"Elif","family":"Yildiz","sequence":"first","affiliation":[{"name":"Harvard Ophthalmology AI Lab, Schepens Eye Research Institute of Massachusetts Eye and Ear, Harvard Medical School, Boston, MA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lucy","family":"Zha","sequence":"additional","affiliation":[{"name":"Harvard Ophthalmology AI Lab, Schepens Eye Research Institute of Massachusetts Eye and Ear, Harvard Medical School, Boston, MA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nazlee","family":"Zebardast","sequence":"additional","affiliation":[{"name":"Harvard Ophthalmology AI Lab, Schepens Eye Research Institute of Massachusetts Eye and Ear, Harvard Medical School, Boston, MA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Min","family":"Shi","sequence":"additional","affiliation":[{"name":"School of Computing and Informatics, University of Louisiana at Lafayette, Lafayette, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mengyu","family":"Wang","sequence":"additional","affiliation":[{"name":"Harvard Ophthalmology AI Lab, Schepens Eye Research Institute of Massachusetts Eye and Ear, Harvard Medical School, Boston, MA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026052913500510000_2026.05.26.26354081v1.1","doi-asserted-by":"publisher","DOI":"10.1001\/jama.2020.21899"},{"key":"2026052913500510000_2026.05.26.26354081v1.2","doi-asserted-by":"publisher","DOI":"10.1016\/j.ophtha.2014.05.013"},{"key":"2026052913500510000_2026.05.26.26354081v1.3","doi-asserted-by":"publisher","DOI":"10.1016\/j.ophtha.2005.07.012"},{"key":"2026052913500510000_2026.05.26.26354081v1.4","doi-asserted-by":"publisher","DOI":"10.1016\/j.ophtha.2006.08.046"},{"key":"2026052913500510000_2026.05.26.26354081v1.5","doi-asserted-by":"publisher","DOI":"10.1001\/archopht.1995.01100050054031"},{"key":"2026052913500510000_2026.05.26.26354081v1.6","doi-asserted-by":"publisher","DOI":"10.1016\/j.ophtha.2007.01.023"},{"key":"2026052913500510000_2026.05.26.26354081v1.7","doi-asserted-by":"publisher","DOI":"10.1001\/archopthalmol.2011.1576"},{"key":"2026052913500510000_2026.05.26.26354081v1.8","doi-asserted-by":"publisher","DOI":"10.1117\/1.jbo.22.12.121604"},{"key":"2026052913500510000_2026.05.26.26354081v1.9","doi-asserted-by":"publisher","DOI":"10.1097\/IJG.0b013e3181af31ec"},{"key":"2026052913500510000_2026.05.26.26354081v1.10","doi-asserted-by":"publisher","DOI":"10.1167\/iovs.14-14630"},{"key":"2026052913500510000_2026.05.26.26354081v1.11","doi-asserted-by":"crossref","first-page":"12","DOI":"10.1016\/j.ajo.2015.09.019","article-title":"Effect of Adjusting Retinal Nerve Fiber Layer Profile to Fovea-Disc Angle Axis on the Thickness and Glaucoma Diagnostic Performance","volume":"161","year":"2016","journal-title":"Am J Ophthalmol"},{"key":"2026052913500510000_2026.05.26.26354081v1.12","doi-asserted-by":"publisher","DOI":"10.1136\/bjophthalmol-2018-312162"},{"key":"2026052913500510000_2026.05.26.26354081v1.13","doi-asserted-by":"publisher","DOI":"10.1167\/iovs.13-12380"},{"key":"2026052913500510000_2026.05.26.26354081v1.14","doi-asserted-by":"publisher","DOI":"10.1001\/archophthalmol.2009.340"},{"key":"2026052913500510000_2026.05.26.26354081v1.15","doi-asserted-by":"publisher","DOI":"10.1097\/IJG.0b013e3182182567"},{"key":"2026052913500510000_2026.05.26.26354081v1.16","doi-asserted-by":"publisher","DOI":"10.1016\/j.ophtha.2018.12.033"},{"issue":"1","key":"2026052913500510000_2026.05.26.26354081v1.17","doi-asserted-by":"crossref","first-page":"4176","DOI":"10.1038\/s41598-021-83720-0","article-title":"Nasal displacement of retinal vessels on the optic disc in glaucoma associated with a nasally angled passage through lamina cribrosa","volume":"11","year":"2021","journal-title":"Sci Rep"},{"issue":"12","key":"2026052913500510000_2026.05.26.26354081v1.18","doi-asserted-by":"crossref","first-page":"1563","DOI":"10.1038\/eye.2012.216","article-title":"Retinal nerve fiber layer and optic disc measurements by spectral domain OCT: normative values and associations in young adults","volume":"26","year":"2012","journal-title":"Eye (Lond)"},{"key":"2026052913500510000_2026.05.26.26354081v1.19","doi-asserted-by":"publisher","DOI":"10.1136\/bjophthalmol-2018-313173"},{"issue":"1","key":"2026052913500510000_2026.05.26.26354081v1.20","doi-asserted-by":"crossref","first-page":"188","DOI":"10.1038\/s41433-020-01191-5","article-title":"Deep learning in glaucoma with optical coherence tomography: a review","volume":"35","year":"2021","journal-title":"Eye (Lond)"},{"issue":"1","key":"2026052913500510000_2026.05.26.26354081v1.21","doi-asserted-by":"crossref","first-page":"558","DOI":"10.1038\/s41598-023-27783-1","article-title":"Assessing the external validity of machine learning-based detection of glaucoma","volume":"13","year":"2023","journal-title":"Sci Rep"},{"issue":"1","key":"2026052913500510000_2026.05.26.26354081v1.22","doi-asserted-by":"crossref","first-page":"48","DOI":"10.1186\/s40942-023-00459-8","article-title":"Fairness and generalizability of OCT normative databases: a comparative analysis","volume":"9","year":"2023","journal-title":"Int J Retina Vitreous"},{"issue":"10","key":"2026052913500510000_2026.05.26.26354081v1.23","doi-asserted-by":"crossref","first-page":"13","DOI":"10.1167\/tvst.12.10.13","article-title":"Normative Percentiles of Retinal Nerve Fiber Layer Thickness and Glaucomatous Visual Field Loss","volume":"12","year":"2023","journal-title":"Transl Vis Sci Technol"},{"key":"2026052913500510000_2026.05.26.26354081v1.24","doi-asserted-by":"crossref","first-page":"2685","DOI":"10.2147\/OPTH.S367722","article-title":"Glaucoma Diagnosis Through the Integration of Optical Coherence Tomography\/Angiography and Machine Learning Diagnostic Models","volume":"16","year":"2022","journal-title":"Clin Ophthalmol"},{"key":"2026052913500510000_2026.05.26.26354081v1.25","doi-asserted-by":"publisher","DOI":"10.1136\/bjo-2023-324792"},{"issue":"7","key":"2026052913500510000_2026.05.26.26354081v1.26","doi-asserted-by":"crossref","first-page":"888","DOI":"10.1016\/j.ophtha.2020.01.023","article-title":"Longitudinal Macular Structure-Function Relationships in Glaucoma","volume":"127","year":"2020","journal-title":"Ophthalmology"},{"issue":"1","key":"2026052913500510000_2026.05.26.26354081v1.27","doi-asserted-by":"crossref","first-page":"95","DOI":"10.1111\/nyas.15043","article-title":"Neurovascular segregation of the retinal nerve fiber layer in glaucoma","volume":"1528","year":"2023","journal-title":"Ann N Y Acad Sci"},{"key":"2026052913500510000_2026.05.26.26354081v1.28","first-page":"832920","article-title":"A Structure-Related Fine-Grained Deep Learning System With Diversity Data for Universal Glaucoma Visual Field Grading","volume":"9","year":"2022","journal-title":"Front Med (Lausanne)"},{"key":"2026052913500510000_2026.05.26.26354081v1.29","doi-asserted-by":"publisher","DOI":"10.1016\/j.ophtha.2009.04.013"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.05.26.26354081","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,29]],"date-time":"2026-05-29T20:50:16Z","timestamp":1780087816000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.64898\/2026.05.26.26354081"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,27]]},"references-count":29,"URL":"https:\/\/doi.org\/10.64898\/2026.05.26.26354081","relation":{},"subject":[],"published":{"date-parts":[[2026,5,27]]},"subtype":"preprint"}}