{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,26]],"date-time":"2026-10-26T19:23:20Z","timestamp":1793042600655,"version":"4.5.0"},"reference-count":24,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,4,13]],"date-time":"2026-04-13T00:00:00Z","timestamp":1776038400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Early detection of keratoconus is essential for preventing postoperative complications in refractive surgery and preserving long-term visual function. Although artificial intelligence has demonstrated strong potential in ophthalmic image analysis, many existing models operate as black-box systems and provide limited clinical interpretability. Transparent decision support is therefore critical for safe deployment of AI in clinical practice.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We propose an explainable neuro-symbolic framework for automated interpretation of corneal topography reports and refractive surgery eligibility assessment. The proposed system integrates multimodal feature extraction, a symbolic corneal knowledge graph, probabilistic reasoning, and large language model (LLM)\u2013based report generation. Quantitative biometric parameters and corneal curvature maps extracted from IOLMaster 700 reports were processed through a hybrid convolutional neural network\u2013Vision Transformer (CNN\u2013ViT) module to capture spatial corneal morphology. These representations were aligned with a clinically curated knowledge graph encoding relationships between corneal parameters, disease states, and surgical decision criteria. Bayesian probabilistic inference was then applied to estimate disease likelihoods, while an ensemble LLM module generated structured bilingual clinical reports explaining the reasoning process.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>In a prospective pilot cohort of 20 eyes, the proposed framework demonstrated strong diagnostic performance for early keratoconus detection, achieving an area under the receiver operating characteristic curve (AUC) of approximately 0.95. Sensitivity and specificity remained high across decision thresholds, and the system achieved a balanced F1 score for refractive surgery eligibility classification. Expert evaluation indicated high interpretability and clinical usefulness of the generated reports. The end-to-end pipeline required approximately 95\u202f\u00b1\u202f12\u202fs per case, supporting near\u2013real-time clinical decision support.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>The proposed neuro-symbolic framework combines deep representation learning, structured medical knowledge, and explainable language-based reporting to provide transparent AI-assisted corneal diagnostics. Although the current results are based on a pilot cohort, the framework demonstrates the potential of integrating neural networks, knowledge graphs, and large language models for interpretable ophthalmic AI systems. Future studies using larger multicenter datasets are required to further validate clinical performance and generalizability.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/frai.2026.1713747","type":"journal-article","created":{"date-parts":[[2026,4,13]],"date-time":"2026-04-13T05:47:33Z","timestamp":1776059253000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":3,"title":["Explainable neuro-symbolic artificial intelligence for automated interpretation of corneal topography and early keratoconus detection"],"prefix":"10.3389","volume":"9","author":[{"given":"Mini Han","family":"Wang","sequence":"first","affiliation":[{"name":"Zhuhai People\u2019s Hospital (The Affiliated Hospital of Beijing Institute of Technology, Zhuhai Clinical Medical College of Jinan University)","place":["Zhuhai, Guangdong, China"]},{"name":"Zhuhai Institute of Advanced Technology Chinese Academy of Sciences","place":["Zhuhai, Guangdong, China"]},{"name":"The Faculty of Medicine, The Chinese University of Hong Kong","place":["Hong Kong, Hong Kong SAR, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shuai","family":"Qin","sequence":"additional","affiliation":[{"name":"Zhuhai People\u2019s Hospital (The Affiliated Hospital of Beijing Institute of Technology, Zhuhai Clinical Medical College of Jinan University)","place":["Zhuhai, Guangdong, China"]},{"name":"Department of Ophthalmology, The Third Affiliated Hospital of Southern Medical University","place":["Guangzhou, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,4,13]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"101072","DOI":"10.1016\/j.xops.2026.101072","article-title":"Hormonal modulation of keratoconus: a systematic review and screening strategy for at-risk populations","volume":"6","author":"Ashraf","year":"2026","journal-title":"Ophthalmol. Sci."},{"key":"ref2","first-page":"180","volume-title":"Neuro-Symbolic AI+ Agent Systems: a First Reflection on Trends, Opportunities and Challenges","author":"Belle","year":"2023"},{"key":"ref3","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1136\/bjo-2024-326097","article-title":"From visual question answering to intelligent AI agents in ophthalmology","volume":"110","author":"Chen","year":"2026","journal-title":"Br. J. Ophthalmol."},{"key":"ref4","doi-asserted-by":"publisher","first-page":"288","DOI":"10.1001\/jamaophthalmol.2025.5844","article-title":"Context for large language model evaluation in ophthalmology","volume":"144","author":"Doherty","year":"2026","journal-title":"JAMA Ophthalmol."},{"key":"ref5","doi-asserted-by":"publisher","first-page":"e0307317","DOI":"10.1371\/journal.pone.0307317","article-title":"A deep learning framework for the early detection of multi-retinal diseases","volume":"19","author":"Ejaz","year":"2024","journal-title":"PLoS One"},{"key":"ref6","doi-asserted-by":"publisher","first-page":"100402","DOI":"10.1016\/j.array.2025.100402","article-title":"Explainable deep learning-based meta-classifier approach for multi-label classification of retinal diseases","volume":"26","author":"Hemal","year":"2025","journal-title":"Array"},{"key":"ref7","doi-asserted-by":"publisher","first-page":"nwac035","DOI":"10.1093\/nsr\/nwac035","article-title":"Neuro-symbolic approaches in artificial intelligence","volume":"9","author":"Hitzler","year":"2022","journal-title":"Natl. Sci. Rev."},{"key":"ref8","doi-asserted-by":"crossref","DOI":"10.3233\/FAIA250239","volume-title":"Neuro-Symbolic AI in Life Sciences","author":"Hoehndorf","year":"2025"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41433-026-04281-y","article-title":"Influence of inflammatory and metabolic factors on keratoconus risk: a causal inference analysis","author":"Hysi","year":"2026","journal-title":"Eye"},{"key":"ref10","doi-asserted-by":"publisher","first-page":"1221","DOI":"10.3390\/diagnostics15101221","article-title":"Diagnostic performance of publicly available large language models in corneal diseases: a comparison with human specialists","volume":"15","author":"Jiao","year":"2025","journal-title":"Diagnostics"},{"key":"ref11","doi-asserted-by":"publisher","first-page":"1252","DOI":"10.1038\/s41433-023-02915-z","article-title":"ChatGPT and beyond: an overview of the growing field of large language models and their use in ophthalmology","volume":"38","author":"Kedia","year":"2024","journal-title":"Eye"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"1411","DOI":"10.1039\/D5SC08461J","article-title":"Explainable artificial intelligence for molecular design in pharmaceutical research","volume":"17","author":"Lamens","year":"2026","journal-title":"Chem. Sci."},{"key":"ref13","doi-asserted-by":"publisher","first-page":"528","DOI":"10.1364\/osac.381120","article-title":"Fast and robust fovea detection framework for OCT images based on foveal avascular zone segmentation","volume":"3","author":"Li","year":"2020","journal-title":"OSA Continuum"},{"key":"ref14","doi-asserted-by":"publisher","first-page":"257","DOI":"10.1007\/s40860-024-00231-1","article-title":"Surveying neuro-symbolic approaches for reliable artificial intelligence of things","volume":"10","author":"Lu","year":"2024","journal-title":"J. Reliab. Intell. Environ."},{"key":"ref15","doi-asserted-by":"publisher","first-page":"4999","DOI":"10.1038\/s41598-025-88948-8","article-title":"Integrated visual and text-based analysis of ophthalmology clinical cases using a large language model","volume":"15","author":"Sorin","year":"2025","journal-title":"Sci. Rep."},{"key":"ref16","doi-asserted-by":"publisher","first-page":"24399","DOI":"10.1038\/s41598-025-09694-5","article-title":"Machine learning-assisted early detection of keratoconus: a comparative analysis of corneal topography and biomechanical data","volume":"15","author":"Syta","year":"2025","journal-title":"Sci. Rep."},{"key":"ref17","doi-asserted-by":"publisher","first-page":"101149","DOI":"10.1016\/j.xops.2026.101149","article-title":"Toward long-term visual field appearance forecasting using artificial intelligence for ophthalmic education and diagnosis","volume":"6","author":"Tian","year":"2026","journal-title":"Ophthalmol. Sci."},{"key":"ref18","doi-asserted-by":"publisher","first-page":"548","DOI":"10.3390\/bioengineering12050548","article-title":"Integrating artificial intelligence and precision therapeutics for advancing the diagnosis and treatment of age-related macular degeneration","volume":"12","author":"Wang","year":"","journal-title":"Bioengineering"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"1174","DOI":"10.3390\/bioengineering12111174","article-title":"An explainable AI framework for corneal imaging interpretation and refractive surgery decision support","volume":"12","author":"Wang","year":"","journal-title":"Bioengineering"},{"key":"ref20","doi-asserted-by":"publisher","first-page":"294","DOI":"10.1002\/oby.70095","article-title":"Artificial intelligence across the obesity continuum: from mechanistic insights to global precision prevention and therapy","volume":"34","author":"Wang","year":"2026","journal-title":"Obesity"},{"key":"ref21","doi-asserted-by":"publisher","first-page":"1527010","DOI":"10.3389\/frai.2025.1527010","article-title":"Applied machine learning in intelligent systems: knowledge graph-enhanced ophthalmic contrastive learning with \u201cclinical profile\u201d prompts","volume":"8","author":"Wang","year":"2025","journal-title":"Front. Artif. Intell."},{"key":"ref22","doi-asserted-by":"publisher","first-page":"445","DOI":"10.26599\/bdma.2023.9020024","article-title":"AI-based advanced approaches and dry eye disease detection based on multi-source evidence: cases, applications, issues, and future directions","volume":"7","author":"Wang","year":"2024","journal-title":"Big Data Min. Anal."},{"key":"ref23","first-page":"470","volume-title":"Fusion Learning Methods for the age-Related Macular Degeneration Diagnosis Based on Multiple Sources of Ophthalmic Digital Images","author":"Wang","year":"2023"},{"key":"ref24","doi-asserted-by":"publisher","first-page":"100967","DOI":"10.1016\/j.landig.2025.100967","article-title":"Agentic artificial intelligence in eye care: is clinical autonomy finally within reach?","volume":"8","author":"Zou","year":"2026","journal-title":"Lancet Digit. Health"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1713747\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,13]],"date-time":"2026-04-13T05:47:35Z","timestamp":1776059255000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1713747\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,4,13]]},"references-count":24,"alternative-id":["10.3389\/frai.2026.1713747"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1713747","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,4,13]]},"article-number":"1713747"}}