EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

AI-Assisted Radiographic Diagnostic Training in Dental Education: A Randomized Pilot Study of Caries Detection and Periodontitis Classification

Selma Dervisbegovic, Susanne Tiefenthaller, Johannes Weimer, Thomas Holzinger, Dino Tur, Christoph Kurzmann, Xiaohui Rausch-Fan, Anita Holzinger, Markus Laky

Applied Sciences · 30.10.2026 · 16 (21) · 10750

DOI: 10.3390/app162110750 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Applied Sciences, online erschienen am 30.10.2026. Verfasst von Selma Dervisbegovic und 8 weiteren Autor:innen. Die Arbeit nennt 26 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Artificial intelligence (AI)-assisted radiographic tools offer new opportunities for diagnostic training in dental education, but evidence regarding their effect on independent student performance remains limited. This prospective randomized pilot study evaluated whether four weeks of supplementary training with an AI-based radiographic diagnostic-support application improved dental students’ performance in caries detection and periodontitis classification compared with conventional learning materials. Of 140 randomized students, 56 were included in the final analysis (AI-assisted group, n = 25; control group, n = 31). Diagnostic performance was assessed using bitewing-based caries tasks and panoramic-radiograph-based periodontitis classification, and outcomes were additionally analyzed according to study phase. The prespecified primary endpoint, total diagnostic performance across caries and periodontal tasks, was not met: scores did not differ significantly between groups (p = 0.251). In secondary analyses, the AI-assisted group achieved numerically higher scores for several outcomes, although the difference in periodontitis extent (localized vs. generalized) was not statistically significant (p = 0.102). In an exploratory adjusted model that was not corrected for multiple comparisons, AI-assisted allocation was associated with better performance for the first caries response position (response position 1; p = 0.026); given the null result for the primary endpoint and the absence of correction for multiple testing, this finding should be regarded as hypothesis-generating rather than confirmatory. Study phase significantly influenced periodontitis grade and caries response position 1. Overall, these findings do not demonstrate a significant benefit of AI-assisted radiographic training on the primary outcome, but the isolated adjusted difference in one caries scoring component supports further evaluation of structured, faculty-supported integration of the AI application into undergraduate dental curricula.

Zitieren

APA

Dervisbegovic, S., Tiefenthaller, S., Weimer, J., Holzinger, T., Tur, D., Kurzmann, C., Rausch-Fan, X., Holzinger, A., & Laky, M. (2026). AI-Assisted Radiographic Diagnostic Training in Dental Education: A Randomized Pilot Study of Caries Detection and Periodontitis Classification. Applied Sciences, 16(21), 10750. https://doi.org/10.3390/app162110750

Vancouver

Dervisbegovic S, Tiefenthaller S, Weimer J, Holzinger T, Tur D, Kurzmann C, et al. AI-Assisted Radiographic Diagnostic Training in Dental Education: A Randomized Pilot Study of Caries Detection and Periodontitis Classification. Applied Sciences. 2026;16(21):10750. doi:10.3390/app162110750

BibTeX
@article{Dervisbegovic2026,
  author = {Dervisbegovic, Selma and Tiefenthaller, Susanne and Weimer, Johannes and Holzinger, Thomas and Tur, Dino and Kurzmann, Christoph and Rausch-Fan, Xiaohui and Holzinger, Anita and Laky, Markus},
  title = {AI-Assisted Radiographic Diagnostic Training in Dental Education: A Randomized Pilot Study of Caries Detection and Periodontitis Classification},
  journal = {Applied Sciences},
  year = {2026},
  volume = {16},
  number = {21},
  pages = {10750},
  doi = {10.3390/app162110750}
}
Kostenlos starten →