{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,31]],"date-time":"2026-08-31T11:16:14Z","timestamp":1788174974903,"version":"build-2803163510"},"reference-count":23,"publisher":"Springer Science and Business Media LLC","license":[{"start":{"date-parts":[[2026,8,31]],"date-time":"2026-08-31T00:00:00Z","timestamp":1788134400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,8,31]],"date-time":"2026-08-31T00:00:00Z","timestamp":1788134400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"name":"Universidade Tuiuti Do Paran\u00e1"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Odontology"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>To evaluate whether off-the-shelf general-purpose multimodal foundation models can achieve clinically meaningful diagnostic performance in detecting secondary caries on interproximal radiographs without domain-specific training, using GPT-5o (OpenAI) and Gemini 2.5 Flash (Google Inc.) as representative publicly accessible models. A comparative exploratory study used 50 interproximal radiographs of posterior teeth obtained from a dental imaging center, yielding 208 teeth. Each tooth was evaluated twice by each model in independent sessions, totaling 416 observations per model. A specialist in restorative dentistry established the reference standard. Models were assessed in a zero-shot setting using a standardized prompt. Diagnostic performance was evaluated at the tooth level using accuracy, sensitivity, specificity, precision, F1-score, and Cohen\u2019s kappa, with 95% confidence intervals estimated via bootstrap resampling (1000 iterations). Both models demonstrated limited diagnostic performance. GPT-5o showed moderate sensitivity (64.7%) but low specificity (32.3%), resulting in an accuracy of 41.3% and negligible agreement with the reference standard (\u03ba\u2009=\u2009\u2009\u2212\u20090.021). Gemini 2.5 Flash exhibited higher sensitivity (82.8%) but extremely low specificity (13.0%), with lower accuracy (32.5%) and similarly negligible agreement (\u03ba\u2009=\u2009\u2009\u2212\u20090.026). Both models showed a strong tendency toward false-positive classifications. Tooth localization performance was poor, with low identification rates in both models. GPT-5o and Gemini 2.5 Flash, when used in their off-the-shelf configurations, demonstrated inadequate diagnostic performance for detecting secondary caries on interproximal radiographs. Their limited accuracy, low specificity, and poor localization ability do not support their direct use for secondary caries detection on interproximal radiographs without domain-specific training and rigorous independent validation.<\/jats:p>","DOI":"10.1007\/s10266-026-01534-w","type":"journal-article","created":{"date-parts":[[2026,8,31]],"date-time":"2026-08-31T10:28:10Z","timestamp":1788172090000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Limitations of off-the-shelf multimodal foundation models in secondary caries detection on interproximal radiographs"],"prefix":"10.1007","author":[{"ORCID":"https:\/\/orcid.org\/0009-0008-2930-5429","authenticated-orcid":false,"given":"Camila Guimar\u00e3es da Costa","family":"Campos","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1325-4248","authenticated-orcid":false,"given":"Cristiano Miranda","family":"de Araujo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-3252-2187","authenticated-orcid":false,"given":"Luana Beatriz das Portas","family":"Luiz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-0337-4206","authenticated-orcid":false,"given":"Diulia Pereira","family":"Bubna","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1852-8423","authenticated-orcid":false,"given":"Angela Graciela Deliga","family":"Schroder","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5649-7234","authenticated-orcid":false,"given":"Flares","family":"Baratto-Filho","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4646-8851","authenticated-orcid":false,"given":"Camila Paiva","family":"Perin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7507-4667","authenticated-orcid":false,"given":"Bianca Marques de Mattos","family":"de Araujo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Thalita","family":"de Paris Matos","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,8,31]]},"reference":[{"key":"1534_CR1","doi-asserted-by":"publisher","first-page":"1869","DOI":"10.1007\/s00784-020-03268-7","volume":"24","author":"H Askar","year":"2020","unstructured":"Askar H, Krois J, G\u00f6stemeyer G, Bottenberg P, Zero D, Banerjee A, et al. Secondary caries: what is it, and how it can be controlled, detected, and managed? Clin Oral Investig. 2020;24:1869\u201376. https:\/\/doi.org\/10.1007\/s00784-020-03268-7.","journal-title":"Clin Oral Investig"},{"key":"1534_CR2","doi-asserted-by":"publisher","first-page":"683","DOI":"10.1007\/s00784-019-02894-0","volume":"24","author":"I Nedeljkovic","year":"2020","unstructured":"Nedeljkovic I, De Munck J, Vanloy A, Declerck D, Lambrechts P, Peumans M, et al. Secondary caries: prevalence, characteristics, and approach. Clin Oral Investig. 2020;24:683\u201391. https:\/\/doi.org\/10.1007\/s00784-019-02894-0.","journal-title":"Clin Oral Investig"},{"key":"1534_CR3","doi-asserted-by":"publisher","first-page":"943","DOI":"10.1177\/0022034514544217","volume":"93","author":"NJ Opdam","year":"2014","unstructured":"Opdam NJ, van de Sande FH, Bronkhorst E, Cenci MS, Bottenberg P, Pallesen U, et al. Longevity of posterior composite restorations: a systematic review and meta-analysis. J Dent Res. 2014;93:943\u20139. https:\/\/doi.org\/10.1177\/0022034514544217.","journal-title":"J Dent Res"},{"key":"1534_CR4","doi-asserted-by":"publisher","first-page":"769","DOI":"10.1177\/0022034520915714","volume":"99","author":"F Schwendicke","year":"2020","unstructured":"Schwendicke F, Samek W, Krois J. Artificial intelligence in dentistry: chances and challenges. J Dent Res. 2020;99:769\u201374. https:\/\/doi.org\/10.1177\/0022034520915714.","journal-title":"J Dent Res"},{"key":"1534_CR5","doi-asserted-by":"publisher","first-page":"117","DOI":"10.1038\/s41432-025-01134-7","volume":"26","author":"SI Mahizha","year":"2025","unstructured":"Mahizha SI, Annrose J, Mano Christaine Angelo J, Domilin Shyni I, Veda Giri GV. Deep convolutional neural networks for early detection of interproximal caries using bitewing radiographs: a systematic review. Evid Based Dent. 2025;26:117. https:\/\/doi.org\/10.1038\/s41432-025-01134-7.","journal-title":"Evid Based Dent"},{"key":"1534_CR6","doi-asserted-by":"publisher","DOI":"10.3390\/jcm9113579","author":"M Prados-Privado","year":"2020","unstructured":"Prados-Privado M, Garc\u00eda Villal\u00f3n J, Mart\u00ednez-Mart\u00ednez CH, Ivorra C, Prados-Frutos JC. Dental caries diagnosis and detection using neural networks: a systematic review. J Clin Med. 2020. https:\/\/doi.org\/10.3390\/jcm9113579.","journal-title":"J Clin Med"},{"key":"1534_CR7","doi-asserted-by":"publisher","first-page":"677","DOI":"10.1007\/s00784-024-06068-5","volume":"28","author":"A Kaur","year":"2024","unstructured":"Kaur A, Jyoti D, Sharma A, Yelam D, Goyal R, Nath A. Deep caries detection using deep learning: from dataset acquisition to detection. Clin Oral Investig. 2024;28:677. https:\/\/doi.org\/10.1007\/s00784-024-06068-5.","journal-title":"Clin Oral Investig"},{"key":"1534_CR8","doi-asserted-by":"publisher","first-page":"1098","DOI":"10.1111\/jerd.13046","volume":"35","author":"F Eggmann","year":"2023","unstructured":"Eggmann F, Weiger R, Zitzmann NU, Blatz MB. Implications of large language models such as ChatGPT for dental medicine. J Esthet Restor Dent. 2023;35:1098\u2013102. https:\/\/doi.org\/10.1111\/jerd.13046.","journal-title":"J Esthet Restor Dent"},{"key":"1534_CR9","doi-asserted-by":"publisher","first-page":"e66207","DOI":"10.2196\/66207","volume":"9","author":"T Zada","year":"2025","unstructured":"Zada T, Tam N, Barnard F, Van Sittert M, Bhat V, Rambhatla S. Medical misinformation in AI-assisted self-diagnosis: development of a method (EvalPrompt) for analyzing large language models. JMIR Form Res. 2025;9:e66207. https:\/\/doi.org\/10.2196\/66207.","journal-title":"JMIR Form Res"},{"key":"1534_CR10","doi-asserted-by":"publisher","DOI":"10.2196\/59505","volume":"26","author":"R AlSaad","year":"2024","unstructured":"AlSaad R, Abd-Alrazaq A, Boughorbel S, Ahmed A, Renault MA, Damseh R, et al. Multimodal large language models in health care: applications, challenges, and future outlook. J Med Internet Res. 2024;26:e59505. https:\/\/doi.org\/10.2196\/59505.","journal-title":"J Med Internet Res"},{"key":"1534_CR11","doi-asserted-by":"publisher","DOI":"10.3389\/froh.2026.1748450","volume":"7","author":"FEA Hassanein","year":"2026","unstructured":"Hassanein FEA, Alkabazi M, Tassoker M, Ahmed Y, Alsaeed S, Abou-Bakr A. Multimodal large language models for oral lesion diagnosis: a systematic review of diagnostic performance and clinical utility. Front Oral Health. 2026;7:1748450. https:\/\/doi.org\/10.3389\/froh.2026.1748450.","journal-title":"Front Oral Health"},{"key":"1534_CR12","doi-asserted-by":"publisher","DOI":"10.1016\/j.ajp.2024.104157","volume":"99","author":"S Thapa","year":"2024","unstructured":"Thapa S, Adhikari S. GPT-4o and multimodal large language models as companions for mental wellbeing. Asian J Psychiatr. 2024;99:104157. https:\/\/doi.org\/10.1016\/j.ajp.2024.104157.","journal-title":"Asian J Psychiatr"},{"key":"1534_CR13","doi-asserted-by":"publisher","DOI":"10.1007\/s00056-025-00624-8","author":"A Abuabara","year":"2025","unstructured":"Abuabara A, Baratto-Filho F, Gallego GR, Luiz L, Meger MN, Scariot R, et al. Comparative evaluation of ChatGPT and Gemini in detecting external apical root resorption on panoramic radiographs of orthodontic patients: artificial intelligence performance in detecting root resorption. J Orofac Orthop. 2025. https:\/\/doi.org\/10.1007\/s00056-025-00624-8.","journal-title":"J Orofac Orthop"},{"key":"1534_CR14","doi-asserted-by":"publisher","DOI":"10.1590\/0103-644020256861","volume":"36","author":"DP Bubna","year":"2026","unstructured":"Bubna DP, Mattos NHR, Luiz L, Baratto-Filho F, Mattos-Calil MT, Silva-Sousa YTC, et al. Can large language models detect periapical lesions in anterior teeth? A comparative study. Braz Dent J. 2026;36:e256861. https:\/\/doi.org\/10.1590\/0103-644020256861.","journal-title":"Braz Dent J"},{"key":"1534_CR15","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.jds.2025.08.016","volume":"21","author":"M Hamada","year":"2026","unstructured":"Hamada M, Kikuchi S, Akitomo T, Kusaka S, Iwamoto Y, Nomura R. Applications and potential of ChatGPT in dentistry: scoping review of research perspectives. J Dent Sci. 2026;21:1\u20138. https:\/\/doi.org\/10.1016\/j.jds.2025.08.016.","journal-title":"J Dent Sci"},{"key":"1534_CR16","doi-asserted-by":"publisher","DOI":"10.1159\/000546448","author":"KSH Horvath","year":"2025","unstructured":"Horvath KSH, Gjerdet NR, Shi X-Q. Advancements in caries diagnostics using bitewing radiography: a systematic review of deep learning approaches. Caries Res. 2025. https:\/\/doi.org\/10.1159\/000546448.","journal-title":"Caries Res"},{"key":"1534_CR17","doi-asserted-by":"publisher","DOI":"10.1186\/s12903-024-04046-7","volume":"24","author":"D Albano","year":"2024","unstructured":"Albano D, Galiano V, Basile M, Di Luca F, Gitto S, Messina C, et al. Artificial intelligence for radiographic imaging detection of caries lesions: a systematic review. BMC Oral Health. 2024;24:274. https:\/\/doi.org\/10.1186\/s12903-024-04046-7.","journal-title":"BMC Oral Health"},{"key":"1534_CR18","doi-asserted-by":"publisher","first-page":"3146","DOI":"10.1007\/s10278-024-01113-x","volume":"37","author":"R Karaku\u015f","year":"2024","unstructured":"Karaku\u015f R, \u00d6zi\u00e7 M, Tassoker M. AI-assisted detection of interproximal, occlusal, and secondary caries on bite-wing radiographs: a single-shot deep learning approach. J Imaging Inform Med. 2024;37:3146\u201359. https:\/\/doi.org\/10.1007\/s10278-024-01113-x.","journal-title":"J Imaging Inform Med"},{"key":"1534_CR19","doi-asserted-by":"publisher","DOI":"10.1016\/j.identj.2025.100848","volume":"75","author":"P Rewthamrongsris","year":"2025","unstructured":"Rewthamrongsris P, Burapacheep J, Phattarataratip E, Kulthanaamondhita P, Tichy A, Schwendicke F, et al. Image-based diagnostic performance of LLMs vs CNNs for oral lichen planus: example-guided and differential diagnosis. Int Dent J. 2025;75:100848. https:\/\/doi.org\/10.1016\/j.identj.2025.100848.","journal-title":"Int Dent J"},{"key":"1534_CR20","doi-asserted-by":"publisher","first-page":"39","DOI":"10.3390\/asi8020039","volume":"8","author":"C Dasanayaka","year":"2025","unstructured":"Dasanayaka C, Dandeniya K, Dissanayake MB, Gunasena C, Jayasinghe R. Multimodal AI and large language models for orthopantomography radiology report generation and Q&A. Appl Syst Innov. 2025;8:39.","journal-title":"Appl Syst Innov"},{"key":"1534_CR21","doi-asserted-by":"publisher","first-page":"179","DOI":"10.1111\/ors.12848","volume":"17","author":"S Tripathy","year":"2024","unstructured":"Tripathy S, Mathur A, Mehta V. A view of neural networks in artificial intelligence in oral pathology. Oral Surg. 2024;17:179\u201380. https:\/\/doi.org\/10.1111\/ors.12848.","journal-title":"Oral Surg"},{"key":"1534_CR22","doi-asserted-by":"publisher","DOI":"10.1002\/cre2.70004","volume":"10","author":"S Negi","year":"2024","unstructured":"Negi S, Mathur A, Tripathy S, Mehta V, Snigdha NT, Adil AH, et al. Artificial intelligence in dental caries diagnosis and detection: an umbrella review. Clin Exp Dent Res. 2024;10:e70004. https:\/\/doi.org\/10.1002\/cre2.70004.","journal-title":"Clin Exp Dent Res"},{"key":"1534_CR23","doi-asserted-by":"publisher","DOI":"10.1016\/j.medp.2024.100030","volume":"1","author":"M Yuan","year":"2024","unstructured":"Yuan M, Bao P, Yuan J, Shen Y, Chen Z, Xie Y, et al. Large language models illuminate a progressive pathway to artificial intelligent healthcare assistant. Med Plus. 2024;1:100030. https:\/\/doi.org\/10.1016\/j.medp.2024.100030.","journal-title":"Med Plus"}],"container-title":["Odontology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10266-026-01534-w.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s10266-026-01534-w","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10266-026-01534-w.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,31]],"date-time":"2026-08-31T10:28:11Z","timestamp":1788172091000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s10266-026-01534-w"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,31]]},"references-count":23,"alternative-id":["1534"],"URL":"https:\/\/doi.org\/10.1007\/s10266-026-01534-w","relation":{},"ISSN":["1618-1247","1618-1255"],"issn-type":[{"value":"1618-1247","type":"print"},{"value":"1618-1255","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,31]]},"assertion":[{"value":"16 April 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"28 July 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"31 August 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors declare that they have no conflict of interest.","order":1,"name":"Ethics","label":"Conflict of interest","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Approved by the Research Ethics Committee (Protocol No. 6202403) in accordance with the Declaration of Helsinki.","order":2,"name":"Ethics","label":"Ethical approval","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Waived due to the retrospective design and use of anonymized data.","order":3,"name":"Ethics","label":"Informed consent","group":{"name":"EthicsHeading","label":"Declarations"}}]}}