{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"Research Square"}],"indexed":{"date-parts":[[2026,6,5]],"date-time":"2026-06-05T11:12:19Z","timestamp":1780657939768,"version":"3.54.1"},"posted":{"date-parts":[[2026,6,5]]},"group-title":"In Review","reference-count":31,"publisher":"Springer Science and Business Media LLC","license":[{"start":{"date-parts":[[2026,6,5]],"date-time":"2026-06-05T00:00:00Z","timestamp":1780617600000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,2,5]]},"abstract":"<title>Abstract<\/title>\n                <p>This work investigated whether deep learning models trained on optic disc-centered fundus images retain sufficient subject-specific information for biometric verification compared with models trained on full-field fundus photographs. A total of 30,836 color fundus photographs from 7,724 eyes of 4,500 subjects were obtained at the Bascom Palmer Eye Institute. Each fundus photograph was processed into three image representations: full-field fundus, optic disc region including 0.5 disc diameters of peripapillary retina, and tightly cropped optic disc only. Images were partitioned at the subject level into training (70%), validation (10%), and test (20%) sets. Separate Siamese convolutional neural network models were trained for each image type using triplet loss to learn subject-discriminative embeddings. Biometric verification was evaluated on the independent test set using exhaustive same-eye image pairing and cosine similarity. All image representations retained measurable subject-specific biometric signal. The full-fundus model achieved the highest performance (AUC, 0.992; EER, 4.4%), followed by the disc-region model (AUC, 0.989; EER, 5.5%) and the disc-only model (AUC, 0.969; EER, 10.5%). Accuracy was 0.968 for full fundus, 0.945 for disc region, and 0.919 for disc-only images. Pairwise comparisons showed significantly worse performance for disc-only images compared with full fundus (P\u2009&lt;\u20090.001). Differences between full-fundus and disc-region models were small and not significant for AUC or EER. These findings demonstrate that deep learning models restricted to optic disc-centered fundus images retain meaningful subject-specific information, although performance declines as available retinal context is reduced. Inclusion of a narrow peripapillary rim yields biometric verification performance comparable to full-field fundus images. Although identity cannot be established from a fundus image without a linking key, recognizing that even restricted retinal regions retain subject-specific features highlights the importance of cautious data-sharing practices while supporting continued scientific collaboration.<\/p>","DOI":"10.21203\/rs.3.rs-8801056\/v1","type":"posted-content","created":{"date-parts":[[2026,6,5]],"date-time":"2026-06-05T10:44:21Z","timestamp":1780656261000},"source":"Crossref","is-referenced-by-count":0,"title":["Optic Disc Fundus Images Retain Biometric Identity Signals Under Deep Learning"],"prefix":"10.21203","author":[{"given":"Ali","family":"Azizi","sequence":"first","affiliation":[{"name":"University of Miami"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rafael","family":"Scherer","sequence":"additional","affiliation":[{"name":"University of Miami"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aaron S.","family":"Rabinowitz","sequence":"additional","affiliation":[{"name":"University of Miami"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Douglas R. da","family":"Costa","sequence":"additional","affiliation":[{"name":"University of Miami"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Vitoria Palazoni Viegas","family":"Mendon\u00e7a","sequence":"additional","affiliation":[{"name":"University of Miami"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Gustavo A.","family":"Samico","sequence":"additional","affiliation":[{"name":"University of Miami"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Gustavo R.","family":"Gameiro","sequence":"additional","affiliation":[{"name":"University of Miami"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Felipe A.","family":"Medeiros","sequence":"additional","affiliation":[{"name":"University of Miami"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","reference":[{"key":"ref1","article-title":"A review: person identification using retinal fundus images","volume":"65","author":"Elangovan P","year":"2019","unstructured":"Elangovan, P. & Nath, M. K. A review: person identification using retinal fundus images. International J. Electron. Telecommunications 65 (2019).","journal-title":"International J. Electron. Telecommunications"},{"key":"ref2","author":"Kachouri R","year":"2020","unstructured":"Kachouri, R., Akil, M. & Elloumi, Y. in Hidden Biometrics: When Biometric Security Meets Biomedical Engineering (ed Amine Nait-ali) 167\u2013184Springer Singapore, (2020)."},{"key":"ref3","doi-asserted-by":"crossref","first-page":"106433","DOI":"10.1016\/j.dib.2020.106433","article-title":"A. RIDB: a dataset of fundus images for retina based person identification","volume":"33","author":"Akram MU","year":"2020","unstructured":"Akram, M. U., Salam, A. A., Khawaja, S. G., Naqvi, S. G. H. & Khan, S. A. RIDB: a dataset of fundus images for retina based person identification. Data Brief. 33, 106433 (2020).","journal-title":"Data Brief."},{"key":"ref4","author":"Devi RM","year":"2022","unstructured":"Devi, R. M. et al. Elsevier,. in Machine Learning for Biometrics 87\u2013104 (2022)."},{"key":"ref5","author":"Shaydyuk NK","year":"2016","unstructured":"Shaydyuk, N. K. & Cleland, T. in IEEE International Carnahan Conference on Security Technology (ICCST). 1\u20135 (IEEE). 1\u20135 (IEEE). (2016)."},{"key":"ref6","doi-asserted-by":"crossref","first-page":"167","DOI":"10.1136\/bjophthalmol-2018-313173","article-title":"Artificial intelligence and deep learning in ophthalmology","volume":"103","author":"Ting DSW","year":"2019","unstructured":"Ting, D. S. W. et al. Artificial intelligence and deep learning in ophthalmology. Br. J. Ophthalmol. 103, 167\u2013175 (2019).","journal-title":"Br. J. Ophthalmol."},{"key":"ref7","first-page":"865","article-title":"Retinal biometric identification using convolutional neural network","volume":"45","author":"Madenda S","year":"2021","unstructured":"Madenda, S., Susetianingtias, D. T., Adlina, D. & Arianty, R. Retinal biometric identification using convolutional neural network. \u041a\u043e\u043c\u043f\u044c\u044e\u0442\u0435\u0440\u043d\u0430\u044f \u043e\u043f\u0442\u0438\u043a\u0430. 45, 865\u2013872 (2021).","journal-title":"\u041a\u043e\u043c\u043f\u044c\u044e\u0442\u0435\u0440\u043d\u0430\u044f \u043e\u043f\u0442\u0438\u043a\u0430"},{"key":"ref8","article-title":"RNFL Thickness in a Population-Based Cohort: The Canadian Longitudinal Study on Aging M2M (Machine-to-Machine) Study","author":"Azizi A","year":"2025","unstructured":"Azizi, A. et al. RNFL Thickness in a Population-Based Cohort: The Canadian Longitudinal Study on Aging M2M (Machine-to-Machine) Study. American J. Ophthalmology (2025).","journal-title":"American J. Ophthalmology"},{"key":"ref9","doi-asserted-by":"crossref","first-page":"383","DOI":"10.1016\/j.ophtha.2020.07.045","article-title":"Detection of progressive glaucomatous optic nerve damage on fundus photographs with deep learning","volume":"128","author":"Medeiros FA","year":"2021","unstructured":"Medeiros, F. A., Jammal, A. A. & Mariottoni, E. B. Detection of progressive glaucomatous optic nerve damage on fundus photographs with deep learning. Ophthalmology 128, 383\u2013392 (2021).","journal-title":"Ophthalmology"},{"key":"ref10","doi-asserted-by":"crossref","first-page":"158","DOI":"10.1038\/s41551-018-0195-0","article-title":"Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning","volume":"2","author":"Poplin R","year":"2018","unstructured":"Poplin, R. et al. Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning. Nat. biomedical Eng. 2, 158\u2013164 (2018).","journal-title":"Nat. biomedical Eng."},{"key":"ref11","doi-asserted-by":"crossref","first-page":"513","DOI":"10.1016\/j.ophtha.2018.12.033","article-title":"From machine to machine: an OCT-trained deep learning algorithm for objective quantification of glaucomatous damage in fundus photographs","volume":"126","author":"Medeiros FA","year":"2019","unstructured":"Medeiros, F. A., Jammal, A. A. & Thompson, A. C. From machine to machine: an OCT-trained deep learning algorithm for objective quantification of glaucomatous damage in fundus photographs. Ophthalmology 126, 513\u2013521 (2019).","journal-title":"Ophthalmology"},{"key":"ref12","doi-asserted-by":"crossref","first-page":"123","DOI":"10.1016\/j.ajo.2019.11.006","article-title":"Human versus machine: comparing a deep learning algorithm to human gradings for detecting glaucoma on fundus photographs","volume":"211","author":"Jammal AA","year":"2020","unstructured":"Jammal, A. A. et al. Human versus machine: comparing a deep learning algorithm to human gradings for detecting glaucoma on fundus photographs. Am. J. Ophthalmol. 211, 123\u2013131 (2020).","journal-title":"Am. J. Ophthalmol."},{"key":"ref13","doi-asserted-by":"crossref","first-page":"1107","DOI":"10.1136\/bjo-2023-324277","article-title":"Prediction of visual field progression with serial optic disc photographs using deep learning","volume":"108","author":"Mohammadzadeh V","year":"2024","unstructured":"Mohammadzadeh, V. et al. Prediction of visual field progression with serial optic disc photographs using deep learning. Br. J. Ophthalmol. 108, 1107\u20131113 (2024).","journal-title":"Br. J. Ophthalmol."},{"key":"ref14","author":"Teh PS","year":"2013","unstructured":"Teh, P. S., Teoh, A. B. J. & Yue, S. A survey of keystroke dynamics biometrics. The Scientific World Journal 408280 (2013). (2013)."},{"key":"ref15","author":"Alliance F","year":"2023","unstructured":"Alliance, F. FIDO Biometric Requirements v3.0 (Final Draft), (2023). https:\/\/fidoalliance.org\/specs\/biometric\/requirements\/Biometrics-Requirements-v3.0-fd-20230111.html"},{"key":"ref16","author":"Fatima J","year":"2013","unstructured":"Fatima, J., Syed, A. M. & Akram, M. U. in IEEE Symposium on Industrial Electronics & Applications. 90\u201395 (IEEE). 90\u201395 (IEEE). (2013)."},{"key":"ref17","doi-asserted-by":"crossref","first-page":"35","DOI":"10.4236\/jsip.2012.31005","article-title":"Retinal identification system based on the combination of fourier and wavelet transform","volume":"3","author":"Sabaghi M","year":"2012","unstructured":"Sabaghi, M., Hadianamrei, S. R., Fattahi, M., Kouchaki, M. R. & Zahedi, A. Retinal identification system based on the combination of fourier and wavelet transform. J. Signal. Inform. Process. 3, 35\u201338 (2012).","journal-title":"J. Signal. Inform. Process."},{"key":"ref18","author":"Dehghani A","year":"2013","unstructured":"Dehghani, A., Ghassabi, Z., Moghddam, H. A. & Moin, M. S. Human recognition based on retinal images and using new similarity function. EURASIP Journal on Image and Video Processing 58 (2013). (2013)."},{"key":"ref19","author":"Waheed Z","year":"2016","unstructured":"Waheed, Z., Waheed, A. & Akram, M. U. in 13th International Bhurban Conference on Applied Sciences and Technology (IBCAST). 101\u2013105 (IEEE). 101\u2013105 (IEEE). (2016)."},{"key":"ref20","doi-asserted-by":"crossref","first-page":"130","DOI":"10.1016\/j.compbiomed.2016.09.019","article-title":"Automated retina identification based on multiscale elastic registration","volume":"79","author":"Figueiredo IN","year":"2016","unstructured":"Figueiredo, I. N. et al. Automated retina identification based on multiscale elastic registration. Comput. Biol. Med. 79, 130\u2013143 (2016).","journal-title":"Comput. Biol. Med."},{"key":"ref21","author":"Akram MU","unstructured":"Akram, M. U., Tariq, A. & Khan, S. A. in 2011 International Conference for Internet Technology and Secured Transactions. 180\u2013184 (IEEE)."},{"key":"ref22","author":"Jiu F","year":"2016","unstructured":"Jiu, F., Noronha, K. & Jayaswal, D. in IEEE 1st international conference on power electronics, intelligent control and energy systems (ICPEICES). 1\u20135 (IEEE). 1\u20135 (IEEE). (2016)."},{"key":"ref23","first-page":"13670","article-title":"A personal identification system using retinal vasculature in retinal fundus images","volume":"38","author":"K\u00f6se C","year":"2011","unstructured":"K\u00f6se, C. & \u0130ki, C. A personal identification system using retinal vasculature in retinal fundus images. Expert Syst. Appl. 38, 13670\u201313681 (2011).","journal-title":"Expert Syst. Appl."},{"key":"ref24","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1155\/2008\/280635","article-title":"A novel retinal identification system","volume":"2008","author":"Farzin H","year":"2008","unstructured":"Farzin, H., Abrishami-Moghaddam, H. & Moin, M. S. A novel retinal identification system. EURASIP J. Adv. Signal Process. 2008, 1\u201310 (2008).","journal-title":"EURASIP J. Adv. Signal Process."},{"key":"ref25","first-page":"191","article-title":"A new approach to biometric person identification from retinal vascular patterns","volume":"3","author":"Gharami RK","year":"2014","unstructured":"Gharami, R. K. & Nath, M. K. A new approach to biometric person identification from retinal vascular patterns. Emerg. Trends Technol. Comput. Sci. (IJETTCS). 3, 191\u2013195 (2014).","journal-title":"Emerg. Trends Technol. Comput. Sci. (IJETTCS)"},{"key":"ref26","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1007\/s11910-025-01417-7","article-title":"The Use Of Fundus Photography In The Emergency Room\u2014A Review","volume":"25","author":"Browning SD","year":"2025","unstructured":"Browning, S. D., Costello, J. M., Dunn, H. P. & Fraser, C. L. The Use Of Fundus Photography In The Emergency Room\u2014A Review. Curr. Neurol. Neurosci. Rep. 25, 1\u201312 (2025).","journal-title":"Curr. Neurol. Neurosci. Rep."},{"key":"ref27","doi-asserted-by":"crossref","first-page":"1694","DOI":"10.1038\/s41433-024-03000-9","article-title":"Head to head comparison of diagnostic performance of three non-mydriatic cameras for diabetic retinopathy screening with artificial intelligence","volume":"38","author":"Do\u011fan ME","year":"2024","unstructured":"Do\u011fan, M. E. et al. Head to head comparison of diagnostic performance of three non-mydriatic cameras for diabetic retinopathy screening with artificial intelligence. Eye 38, 1694\u20131701 (2024).","journal-title":"Eye"},{"key":"ref28","doi-asserted-by":"crossref","first-page":"1543","DOI":"10.3390\/diagnostics14141543","article-title":"The assessment of retinal image quality using a non-mydriatic fundus camera in a teleophthalmologic platform","volume":"14","author":"Chan TY","year":"2024","unstructured":"Chan, T. Y., Wang, J. H., Chen, N. & Chiu, C. J. The assessment of retinal image quality using a non-mydriatic fundus camera in a teleophthalmologic platform. Diagnostics 14, 1543 (2024).","journal-title":"Diagnostics"},{"key":"ref29","article-title":"Balancing benefits and risks: the case for retinal images to be considered as non-protected health information for research purposes","volume":"10","author":"Lum F","year":"2024","unstructured":"Lum, F. Balancing benefits and risks: the case for retinal images to be considered as non-protected health information for research purposes. Ophthalmology 10 (2024).","journal-title":"Ophthalmology"},{"key":"ref30","doi-asserted-by":"crossref","first-page":"110341","DOI":"10.1016\/j.exer.2025.110341","article-title":"Retinal imaging in an era of open science and privacy protection","volume":"255","author":"Gim N","year":"2025","unstructured":"Gim, N. et al. Retinal imaging in an era of open science and privacy protection. Exp. Eye Res. 255, 110341 (2025).","journal-title":"Exp. Eye Res."},{"key":"ref31","doi-asserted-by":"crossref","first-page":"5466","DOI":"10.1364\/BOE.495438","article-title":"Unobtrusive biometric data de-identification of fundus images using latent space disentanglement","volume":"14","author":"Zhao Z","year":"2023","unstructured":"Zhao, Z. et al. Unobtrusive biometric data de-identification of fundus images using latent space disentanglement. Biomedical Opt. Express. 14, 5466\u20135483 (2023).","journal-title":"Biomedical Opt. Express"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/www.researchsquare.com\/article\/rs-8801056\/v1","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.researchsquare.com\/article\/rs-8801056\/v1.html","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,5]],"date-time":"2026-06-05T10:45:39Z","timestamp":1780656339000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.researchsquare.com\/article\/rs-8801056\/v1"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,5]]},"references-count":31,"URL":"https:\/\/doi.org\/10.21203\/rs.3.rs-8801056\/v1","relation":{},"subject":[],"published":{"date-parts":[[2026,6,5]]},"subtype":"preprint"}}