EUVIMEDEUVIMED entdecken
Crossref

Caries Detection in Primary Molars with Bitewing Radiographs through Deep Learning Based-Object Detectors

Zhi Qin Tan, Ilana Felix Pinho, Ryan Banks, Xiongjie Chen, Laura Regina Antunes Pontes, Mariana Minatel Braga, Fausto Medeiros Mendes, Yunpeng Li

Caries Research · 18.12.2025 · 1-9

DOI: 10.1159/000550079 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Caries Research, online erschienen am 18.12.2025. Verfasst von Zhi Qin Tan und 7 weiteren Autor:innen. Beteiligte Einrichtung: Faculty of Dentistry, Oral and Craniofacial Sciences, King's College London, London, UK. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 28 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

Automated methods for caries detection among children can help triage children with greater need for treatment and assist clinicians in achieving a more accurate diagnosis of caries lesions. Therefore, we aimed to develop artificial intelligence algorithms based on various object detectors for detecting and staging caries lesions in primary molars using bitewing radiographs.

Methods

We used a dataset of 1,023 bitewing radiographs of primary molars from children aged 3–10 years. The radiographs were annotated by two examiners using a dedicated annotation platform to draw bounding boxes corresponding to four different primary caries severity stages. Five popular deep learning object detection algorithms were trained to detect the annotated caries and evaluated with various performance metrics, considering three different thresholds: all lesions, dentine caries lesions, and dentine caries lesions requiring operative treatment.

Results

For staging caries lesions, the DINO model achieved higher concordance with a weighted kappa score of 0.713 on the test dataset, outperforming other object detectors. The DINO model achieved the highest sensitivity for detecting all caries lesions (0.509) while also attaining the highest sensitivity and accuracy in identifying dentine caries requiring operative treatment, at 0.659 and 0.980, respectively. However, YOLOv7 also achieved good performance, with specificity values exceeding 0.99 and accuracy values exceeding 0.92 across all thresholds.

Conclusions

The DINO and YOLOv7 algorithms perform well at detecting caries on primary molars in bitewings, highlighting their potential for clinical application at aiding clinicians in daily practice.

Schlagwörter

Stichwörter der Autor:innen

Artificial intelligenceBitewingCaries detectionDeep learningPrimary teeth

Autor:innen und Institutionen

  • Faculty of Dentistry, Oral and Craniofacial Sciences, King's College London, London, UK.
  • Computer Science Research Centre, University of Surrey, Guildford, UK.
  • Department of Pediatric Dentistry, School of Dentistry, University of São Paulo, Sao Paulo, Brazil.
  • Department of Pediatric Dentistry, School of Dentistry, University of São Paulo, Sao Paulo, Brazil, fmmendes@usp.br.
  • Faculty of Dentistry, Oral and Craniofacial Sciences, King's College London, London, UK, yunpeng.li@kcl.ac.uk.
  • Computer Science Research Centre, University of Surrey, Guildford, UK, yunpeng.li@kcl.ac.uk.

Zitieren

APA

Tan, Z. Q., Pinho, I. F., Banks, R., Chen, X., Pontes, L. R. A., Braga, M. M., Mendes, F. M., & Li, Y. (2025). Caries Detection in Primary Molars with Bitewing Radiographs through Deep Learning Based-Object Detectors. Caries Research, 1-9. https://doi.org/10.1159/000550079

Vancouver

Tan ZQ, Pinho IF, Banks R, Chen X, Pontes LRA, Braga MM, et al. Caries Detection in Primary Molars with Bitewing Radiographs through Deep Learning Based-Object Detectors. Caries Research. 2025:1-9. doi:10.1159/000550079

BibTeX
@article{Tan2025,
  author = {Tan, Zhi Qin and Pinho, Ilana Felix and Banks, Ryan and Chen, Xiongjie and Pontes, Laura Regina Antunes and Braga, Mariana Minatel and Mendes, Fausto Medeiros and Li, Yunpeng},
  title = {Caries Detection in Primary Molars with Bitewing Radiographs through Deep Learning Based-Object Detectors},
  journal = {Caries Research},
  year = {2025},
  pages = {1--9},
  doi = {10.1159/000550079}
}
Kostenlos starten →