{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,26]],"date-time":"2026-10-26T19:24:21Z","timestamp":1793042661694,"version":"4.5.0"},"reference-count":41,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2025,3,20]],"date-time":"2025-03-20T00:00:00Z","timestamp":1742428800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2025,3,20]],"date-time":"2025-03-20T00:00:00Z","timestamp":1742428800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100009594","name":"University of P\u00e9cs","doi-asserted-by":"crossref","id":[{"id":"10.13039\/501100009594","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Sci Rep"],"abstract":"<jats:title>Abstract<\/jats:title>\n          <jats:p>To provide an artificial intelligence (AI) method using in vivo confocal microscopy (IVCM) to differentiate ocular surface squamous neoplasia (OSSN) from other lesions and compare the performance of well-known AI-related solutions. A dataset of 2,774 IVCM images, comprising OSSN and other ocular surface diseases was used to train three deep learning models: ResNet50V2, Yolov8x, and VGG19. These models were trained to identify OSSN-related lesions by recognizing specific visual features, including the \u201cstarry-sky\u201d pattern, hyperkeratosis, mitotic figures and irregularly shaped epithelial cells. To mitigate class imbalance, a novel square-based data augmentation strategy was employed. Additionally, we implemented a few-shot learning model to enhance the precision of rare symptoms, such as mitosis. To enhance model interpretation, Shapley values and Uniform Manifold Approximation and Projection (UMAP) analysis were employed to explain decision-making processes. The AI models demonstrated high accuracy in distinguishing healthy tissues from pathological ones, achieving over 90% accuracy across all models. In our binary classification task, all AI models had accuracy above 97% (precision\u2009\u2265\u200998%, recall\u2009\u2265\u200985%, F1 score\u2009\u2265\u200992%). The model achieved lower accuracy in 4 class labeled classification. Aggregation of cell-level results provided the best performance with an F1 score of 100%. The models successfully identified patient-specific features in IVCM images, suggesting that these images can act as \u201cfingerprints\u201d. Our AI model utilizing IVCM was able to classify OSSN with high accuracy. Moreover, cell-level classification results could be backpropagated to image-level and patient-level. The patient-specific information within IVCM images offers promise for personalized diagnostics and treatment monitoring in ocular oncology.<\/jats:p>","DOI":"10.1038\/s41598-025-94876-4","type":"journal-article","created":{"date-parts":[[2025,3,20]],"date-time":"2025-03-20T01:32:36Z","timestamp":1742434356000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":20,"title":["Artificial intelligence to enhance the diagnosis of ocular surface squamous neoplasia"],"prefix":"10.1038","volume":"15","author":[{"given":"Kincs\u0151","family":"Kozma","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zolt\u00e1n Rich\u00e1rd","family":"J\u00e1nki","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Vilmos","family":"Bilicki","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Adrienne","family":"Csutak","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Eszter","family":"Szalai","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2025,3,20]]},"reference":[{"key":"94876_CR1","doi-asserted-by":"publisher","first-page":"172","DOI":"10.1016\/j.exer.2014.10.015","volume":"129","author":"S Gichuhi","year":"2014","unstructured":"Gichuhi, S., Ohnuma, S., Sagoo, M. S. & Burton, M. J. Pathophysiology of ocular surface squamous neoplasia. Exp. Eye Res. 129, 172\u2013182. https:\/\/doi.org\/10.1016\/j.exer.2014.10.015 (2014).","journal-title":"Exp. Eye Res."},{"issue":"1","key":"94876_CR2","doi-asserted-by":"publisher","first-page":"14221","DOI":"10.1038\/s41598-022-18545-6","volume":"12","author":"K Kozma","year":"2022","unstructured":"Kozma, K. et al. Topical pharmacotherapy for ocular surface squamous neoplasia: systematic review and meta-analysis. Sci. Rep. 12 (1), 14221. https:\/\/doi.org\/10.1038\/s41598-022-18545-6 (2022).","journal-title":"Sci. Rep."},{"key":"94876_CR3","doi-asserted-by":"publisher","unstructured":"Ong, S. S., Vora, G. K. & Gupta, P. K. Anterior segment imaging in ocular surface squamous neoplasia. J. Ophthalmol. 2016 (5435092). https:\/\/doi.org\/10.1155\/2016\/5435092 (2016).","DOI":"10.1155\/2016\/5435092"},{"issue":"3","key":"94876_CR4","doi-asserted-by":"publisher","first-page":"334","DOI":"10.1111\/ceo.13512","volume":"47","author":"DV Patel","year":"2019","unstructured":"Patel, D. V., Zhang, J. & McGhee, C. N. In vivo confocal microscopy of the inflamed anterior segment: A review of clinical and research applications. Clin. Exp. Ophthalmol. 47 (3), 334\u2013345. https:\/\/doi.org\/10.1111\/ceo.13512 (2019).","journal-title":"Clin. Exp. Ophthalmol."},{"issue":"2","key":"94876_CR5","doi-asserted-by":"publisher","first-page":"167","DOI":"10.1136\/bjophthalmol-2018-313173","volume":"103","author":"DSW Ting","year":"2019","unstructured":"Ting, D. S. W. et al. Artificial intelligence and deep learning in ophthalmology. Br. J. Ophthalmol. 103 (2), 167\u2013175. https:\/\/doi.org\/10.1136\/bjophthalmol-2018-313173 (2019).","journal-title":"Br. J. Ophthalmol."},{"issue":"5","key":"94876_CR6","doi-asserted-by":"publisher","first-page":"1122","DOI":"10.1016\/j.cell.2018.02.010","volume":"172","author":"DS Kermany","year":"2018","unstructured":"Kermany, D. S. et al. Identifying medical diagnoses and treatable diseases by Image-Based deep learning. Cell 172 (5), 1122\u20131131e9. https:\/\/doi.org\/10.1016\/j.cell.2018.02.010 (2018).","journal-title":"Cell"},{"issue":"2","key":"94876_CR7","doi-asserted-by":"publisher","first-page":"109","DOI":"10.22608\/APO.201710","volume":"6","author":"CL Shields","year":"2017","unstructured":"Shields, C. L. et al. Conjunctival tumors: review of clinical features, risks, biomarkers, and Outcomes\u2013The 2017 J. Donald M. Gass lecture. Asia Pac. J. Ophthalmol. (Phila). 6 (2), 109\u2013120. https:\/\/doi.org\/10.22608\/APO.201710 (2017).","journal-title":"Asia Pac. J. Ophthalmol. (Phila)"},{"key":"94876_CR8","doi-asserted-by":"publisher","first-page":"104418","DOI":"10.1016\/j.compbiomed.2021.104418","volume":"135","author":"M Abdar","year":"2021","unstructured":"Abdar, M. et al. Uncertainty quantification in skin cancer classification using three-way decision-based bayesian deep learning. Comput. Biol. Med. 135, 104418. https:\/\/doi.org\/10.1016\/j.compbiomed.2021.104418 (2021).","journal-title":"Comput. Biol. Med."},{"key":"94876_CR9","doi-asserted-by":"publisher","first-page":"484","DOI":"10.1016\/j.ophtha.2013.09.027","volume":"121.2","author":"MB Nguena","year":"2014","unstructured":"Nguena, M. B. et al. Diagnosing ocular surface squamous neoplasia in East Africa: case-control study of clinical and in vivo confocal microscopy assessment. Ophthalmology 121.2, 484\u2013491. https:\/\/doi.org\/10.1016\/j.ophtha.2013.09.027 (2014).","journal-title":"Ophthalmology"},{"issue":"6","key":"94876_CR10","doi-asserted-by":"publisher","first-page":"781","DOI":"10.1038\/eye.2012.15","volume":"26","author":"Y Xu","year":"2012","unstructured":"Xu, Y. et al. The clinical value of in vivo confocal microscopy for diagnosis of ocular surface squamous neoplasia. Eye (Lond). 26 (6), 781\u2013787. https:\/\/doi.org\/10.1038\/eye.2012.15 (2012).","journal-title":"Eye (Lond)"},{"key":"94876_CR11","unstructured":"Sekachev, B., Zhavoronkov, A. & Manovich, N. Computer vision annotation tool: a universal approach to data annotation. (2019). https:\/\/github.com\/openvinotoolkit\/cvat"},{"key":"94876_CR12","unstructured":"Jocher, G., Chaurasia, A., Qiu, J. & Ultralytics, Y. O. L. O. (Version 8.0.0) [Software] by Ultralytics. (2023). https:\/\/github.com\/ultralytics\/ultralytics"},{"key":"94876_CR13","doi-asserted-by":"crossref","unstructured":"He, K., Zhang, X., Ren, S. & Sun, J. Identity mappings in deep residual networks in Computer Vision\u2013ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11\u201314, Proceedings, Part IV 14 630\u2013645 (Springer International Publishing 2016). (2016).","DOI":"10.1007\/978-3-319-46493-0_38"},{"key":"94876_CR14","unstructured":"Simonyan, K. & Zisserman, A. Very deep convolutional networks for large-scale image recognition. Preprint at (2014). https:\/\/arxiv.org\/abs\/1409.1556"},{"key":"94876_CR15","unstructured":"Bennequin, E. easyfsl [Software]. Universit\u00e9 Paris-Saclay. (2022). https:\/\/github.com\/sicara\/easy-few-shot-learning"},{"key":"94876_CR16","doi-asserted-by":"publisher","first-page":"42","DOI":"10.1186\/s13073-021-00845-7","volume":"13","author":"H Chereda","year":"2021","unstructured":"Chereda, H. et al. Explaining decisions of graph convolutional neural networks: patient-specific molecular subnetworks responsible for metastasis prediction in breast cancer. Genome Med. 13, 42. https:\/\/doi.org\/10.1186\/s13073-021-00845-7 (2021).","journal-title":"Genome Med."},{"key":"94876_CR17","doi-asserted-by":"publisher","unstructured":"Kiranyaz, S., Ince, T., Hamila, R. & Gabbouj, M. Convolutional Neural Networks for patient-specific ECG classification. 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). Milan, Italy. 2608\u20132611; (2015). https:\/\/doi.org\/10.1109\/EMBC.2015.7318926 (2015).","DOI":"10.1109\/EMBC.2015.7318926"},{"key":"94876_CR18","doi-asserted-by":"crossref","unstructured":"Kirillov, A. et al. Segment anything in Proceedings of the IEEE\/CVF International Conference on Computer Vision 4015\u20134026 (2023).","DOI":"10.1109\/ICCV51070.2023.00371"},{"key":"94876_CR19","doi-asserted-by":"publisher","DOI":"10.5281\/zenodo.12776091","author":"iMED (Ningbo Institute of Materials Technology and Engineering","year":"2024","unstructured":"iMED (Ningbo Institute of Materials Technology and Engineering. Chinese academy of sciences, Ningbo, China), \u201eCORN: corneal nerve fiber dataset. Zenodo https:\/\/doi.org\/10.5281\/zenodo.12776091 (2024).","journal-title":"Zenodo"},{"key":"94876_CR20","doi-asserted-by":"publisher","first-page":"106584","DOI":"10.1016\/j.cmpb.2021.106584","volume":"214","author":"Y Nohara","year":"2022","unstructured":"Nohara, Y., Matsumoto, K., Soejima, H. & Nakashima, N. Explanation of machine learning models using Shapley additive explanation and application for real data in hospital. Comput. Methods Programs Biomed. 214, 106584. https:\/\/doi.org\/10.1016\/j.cmpb.2021.106584 (2022).","journal-title":"Comput. Methods Programs Biomed."},{"key":"94876_CR21","unstructured":"Lundberg, S. M. & Lee, S. I. A unified approach to interpreting model predictions. Advances in neural information processing systems, 30. Prepint at (2017). https:\/\/arxiv.org\/abs\/1705.07874"},{"key":"94876_CR22","unstructured":"Lundberg, S. M. SHAP [Software], (2017). https:\/\/shap.readthedocs.io\/en\/latest\/index.html"},{"key":"94876_CR23","doi-asserted-by":"crossref","unstructured":"McInnes, L. & Healy, J. UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction. Prepint at (2018). https:\/\/arxiv.org\/abs\/1802.03426https:\/\/umap-learn.readthedocs.io\/en\/latest\/.","DOI":"10.21105\/joss.00861"},{"issue":"2","key":"94876_CR24","doi-asserted-by":"publisher","first-page":"147","DOI":"10.1002\/j.1538-7305.1950.tb00463.x","volume":"29","author":"RW Hamming","year":"1950","unstructured":"Hamming, R. W. Error detecting and error correcting codes. Bell Syst. Tech. J. 29 (2), 147\u2013160 (1950).","journal-title":"Bell Syst. Tech. J."},{"key":"94876_CR25","first-page":"547","volume":"37","author":"P Yaccard","year":"1901","unstructured":"Yaccard, P. Etude comparative de La distribution Florale Dans Une portion des Alpes et des Jura. Bull. Del. La. Soci\u00b4et\u00b4e Vaudoise Des. Sci. Nat. 37, 547\u2013579 (1901).","journal-title":"Bull. Del. La. Soci\u00b4et\u00b4e Vaudoise Des. Sci. Nat."},{"key":"94876_CR26","doi-asserted-by":"publisher","first-page":"257","DOI":"10.1098\/rsta.1900.0019","volume":"194","author":"GU Yule","year":"1900","unstructured":"Yule, G. U. On the association of attributes in statistics: with illustrations from the material of the childhood society, &c. Philosophical Trans. Royal Soc. Lond. Ser. A. 194, 257\u2013319 (1900).","journal-title":"Philosophical Trans. Royal Soc. Lond. Ser. A"},{"key":"94876_CR27","first-page":"193","volume":"4","author":"PF Russell","year":"1940","unstructured":"Russell, P. F. & Rao, T. R. On habitat and association of species of anophelines as shown by the use of square tables. Indian J. Malariol. 4, 193\u2013197 (1940).","journal-title":"Indian J. Malariol."},{"key":"94876_CR28","doi-asserted-by":"publisher","first-page":"336","DOI":"10.1007\/s11263-019-01228-7","volume":"128","author":"RR Selvaraju","year":"2020","unstructured":"Selvaraju, R. R. Grad-CAM: visual explanations from deep networks via gradient-based localization. Int. J. Comput. Vis. 128, 336\u2013359 (2020).","journal-title":"Int. J. Comput. Vis."},{"issue":"10","key":"94876_CR29","doi-asserted-by":"publisher","first-page":"1340","DOI":"10.1093\/bioinformatics\/btq134","volume":"26","author":"A Altmann","year":"2010","unstructured":"Altmann, A., Tolosi, L., Sander, O. & Lengauer, T. Permutation importance: a corrected feature importance measure. Bioinformatics 26 (10), 1340\u20131347. https:\/\/doi.org\/10.1093\/bioinformatics\/btq134 (2010).","journal-title":"Bioinformatics"},{"issue":"12","key":"94876_CR30","doi-asserted-by":"publisher","first-page":"1728","DOI":"10.1136\/bjophthalmol-2017-311530","volume":"102","author":"LA Dalvin","year":"2018","unstructured":"Dalvin, L. A., Salom\u00e3o, D. R. & Patel, S. V. Population-based incidence of conjunctival tumours in olmsted County, Minnesota. Br. J. Ophthalmol. 102 (12), 1728\u20131734. https:\/\/doi.org\/10.1136\/bjophthalmol-2017-311530 (2018).","journal-title":"Br. J. Ophthalmol."},{"issue":"1","key":"94876_CR31","doi-asserted-by":"publisher","first-page":"11","DOI":"10.1080\/17469899.2017.1263567","volume":"12","author":"IO Sayed-Ahmed","year":"2017","unstructured":"Sayed-Ahmed, I. O., Palioura, S., Galor, A. & Karp, C. L. Diagnosis and medical management of ocular surface squamous neoplasia. Expert Rev. Ophthalmol. 12 (1), 11\u201319. https:\/\/doi.org\/10.1080\/17469899.2017.1263567 (2017).","journal-title":"Expert Rev. Ophthalmol."},{"issue":"4","key":"94876_CR32","doi-asserted-by":"publisher","first-page":"369","DOI":"10.1097\/ICU.0000000000000771","volume":"32","author":"N Venkateswaran","year":"2021","unstructured":"Venkateswaran, N., Sripawadkul, W. & Karp, C. L. The role of imaging technologies for ocular surface tumors. Curr. Opin. Ophthalmol. 32 (4), 369\u2013378. https:\/\/doi.org\/10.1097\/ICU.0000000000000771 (2021).","journal-title":"Curr. Opin. Ophthalmol."},{"issue":"7","key":"94876_CR33","doi-asserted-by":"publisher","first-page":"694","DOI":"10.3390\/diagnostics14070694","volume":"14","author":"K Kryszan","year":"2024","unstructured":"Kryszan, K. et al. Artificial-Intelligence-Enhanced analysis of in vivo confocal microscopy in corneal diseases: A review. Diagnostics (Basel). 14 (7), 694. https:\/\/doi.org\/10.3390\/diagnostics14070694 (2024).","journal-title":"Diagnostics (Basel)"},{"issue":"6","key":"94876_CR34","doi-asserted-by":"publisher","first-page":"729","DOI":"10.1097\/ICO.0000000000002887","volume":"41","author":"V Aykut","year":"2022","unstructured":"Aykut, V. et al. Corneal sensation and nerve fiber changes in patients with interstitial cystitis. Cornea 41 (6), 729\u2013733. https:\/\/doi.org\/10.1097\/ICO.0000000000002887 (2022).","journal-title":"Cornea"},{"issue":"2","key":"94876_CR35","doi-asserted-by":"publisher","first-page":"235","DOI":"10.1016\/j.jtos.2017.12.003","volume":"16","author":"M Zarei-Ghanavati","year":"2018","unstructured":"Zarei-Ghanavati, M. et al. Changes in in vivo confocal microscopic findings of ocular surface squamous neoplasia during treatment with topical interferon alfa-2b. Ocul Surf. 16 (2), 235\u2013241. https:\/\/doi.org\/10.1016\/j.jtos.2017.12.003 (2018).","journal-title":"Ocul Surf."},{"key":"94876_CR36","unstructured":"Bennequin, E. (n.d.). easyfsl [Software]. Universit\u00e9 Paris-Saclay. (2022). https:\/\/github.com\/sicara\/easy-few-shot-learning"},{"issue":"1","key":"94876_CR37","doi-asserted-by":"publisher","first-page":"8366","DOI":"10.1038\/s41598-021-87762-2","volume":"11","author":"S Tang","year":"2021","unstructured":"Tang, S. et al. Data valuation for medical imaging using Shapley value and application to a large-scale chest X-ray dataset. Sci. Rep. 11 (1), 8366. https:\/\/doi.org\/10.1038\/s41598-021-87762-2 (2021).","journal-title":"Sci. Rep."},{"key":"94876_CR38","doi-asserted-by":"publisher","DOI":"10.1038\/nbt.4314","author":"E Becht","year":"2018","unstructured":"Becht, E. et al. Dimensionality reduction for visualizing single-cell data using UMAP. Nat. Biotechnol. https:\/\/doi.org\/10.1038\/nbt.4314 (2018).","journal-title":"Nat. Biotechnol."},{"issue":"2","key":"94876_CR39","doi-asserted-by":"publisher","first-page":"156","DOI":"10.1038\/s41587-020-00809-z","volume":"39","author":"D Kobak","year":"2021","unstructured":"Kobak, D. & Linderman, G. C. Initialization is critical for preserving global data structure in both t-SNE and UMAP. Nat. Biotechnol. 39 (2), 156\u2013157. https:\/\/doi.org\/10.1038\/s41587-020-00809-z (2021).","journal-title":"Nat. Biotechnol."},{"issue":"7","key":"94876_CR40","doi-asserted-by":"publisher","first-page":"795","DOI":"10.1016\/j.amjmed.2019.01.017","volume":"132","author":"N Noorbakhsh-Sabet","year":"2019","unstructured":"Noorbakhsh-Sabet, N., Zand, R., Zhang, Y. & Abedi, V. Artificial intelligence transforms the future of health care. Am. J. Med. 132 (7), 795\u2013780. https:\/\/doi.org\/10.1016\/j.amjmed.2019.01.017 (2019).","journal-title":"Am. J. Med."},{"issue":"1","key":"94876_CR41","doi-asserted-by":"publisher","first-page":"44","DOI":"10.1038\/s41591-018-0300-7","volume":"25","author":"EJ Topol","year":"2019","unstructured":"Topol, E. J. High-performance medicine: the convergence of human and artificial intelligence. Nat. Med. 25 (1), 44\u201356. https:\/\/doi.org\/10.1038\/s41591-018-0300-7 (2019).","journal-title":"Nat. Med."}],"container-title":["Scientific Reports"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41598-025-94876-4.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-025-94876-4","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-025-94876-4.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,3,20]],"date-time":"2025-03-20T01:32:57Z","timestamp":1742434377000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41598-025-94876-4"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,3,20]]},"references-count":41,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2025,12]]}},"alternative-id":["94876"],"URL":"https:\/\/doi.org\/10.1038\/s41598-025-94876-4","relation":{},"ISSN":["2045-2322"],"issn-type":[{"value":"2045-2322","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,3,20]]},"assertion":[{"value":"22 September 2024","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"17 March 2025","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"20 March 2025","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare no competing interests.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}},{"value":"Not required as data is not individualized and primary data was not collected.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethics approval and consent to participate"}},{"value":"The corresponding author accepts responsibility for releasing this material on behalf of all co-authors","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for publication"}}],"article-number":"9550"}}