EUVIMEDEUVIMED entdecken
ValidierungsstudieOpen Access · CC BY-NC 4.0

AI ‐Driven Detection and Measurement of Keratinized Gingiva in Dental Photographs: Validation Using Reference Retainers

Ya-Chi Chen iD, Ling Chen iD, Yu-Lin Lai iD, Wei-Ting Chang iD, Shyh-Yuan Lee iD

Journal of Clinical Periodontology · 2025

DOI: 10.1111/jcpe.14164 ↗

Kurz gefasst

Validierungsstudie in der Zeitschrift Journal of Clinical Periodontology, 2025. Verfasst von Ya-Chi Chen und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Stomatology, Taipei Veterans General Hospital, Taipei, Taiwan. Hauptthemen laut MEDLINE: Gingiva, Deep Learning, Photography, Dental. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 51 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: National Science and Technology Council.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Aim

To evaluate a deep learning (DL) model for detecting keratinized gingiva (KG) in dental photographs and validate its clinical applicability using reference retainers for calibration. Materials and Methods A total of 576 sextant photographs were selected from 32 subjects, each with three sets of photographs: iodine‐stained, unstained and line‐marked retainers. Relative keratinized gingiva width (rKGW) was measured using visual, functional and histochemical staining methods with reference retainers. A pre‐trained DeepLabv3 model with ResNet50 backbone was fine‐tuned to predict KG areas, which were then applied to the photographs with line‐marked retainers for subsequent rKGW measurement.

Results

The AI model achieved a Dice coefficient of 93.30% and an accuracy of 93.32%. Using histochemical measurements as gold standards, the absolute differences in rKGW of AI measurements were statistically insignificant with visual ( p = 0.935) and functional ( p = 0.979) measurements. The adjusted difference between AI and histochemical measurements was 0.377 mm. AI closely matched histochemical measurements in the maxillary anterior region (0.011 mm, p = 0.903) but was significantly higher in the maxillary posterior region (0.327 mm, p < 0.05).

Conclusions

The proposed AI model is the first to reliably identify full‐mouth KG, validated thoroughly using reference retainers. However, predictions for posterior teeth warrant further improvement.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

GingivaDeep LearningPhotography, DentalHumansMaleFemaleAdultMiddle Aged

Stichwörter der Autor:innen

artificial intelligencedeep learninggingivaphotographretainers

Autor:innen und Institutionen

  • Department of Stomatology, Taipei Veterans General Hospital, Taipei, Taiwan.
  • School of Dentistry, National Yang Ming Chiao Tung University, Taipei, Taiwan.
  • Institute of Hospital and Health Care Administration, National Yang Ming Chiao Tung University, Taipei, Taiwan.
  • Department of Dentistry, Taipei City Hospital, Taipei, Taiwan.
  • Oral Medicine Innovation Center, National Yang Ming Chiao Tung University, Taipei, Taiwan.

ORCID-Profile: Ya-Chi Chen, Ling Chen, Yu-Lin Lai, Wei-Ting Chang, Shyh-Yuan Lee

Förderung

  • National Science and Technology Council

Zitieren

APA

Chen, Y. C., Chen, L., Lai, Y. L., Chang, W. T., & Lee, S. Y. (2025). AI ‐Driven Detection and Measurement of Keratinized Gingiva in Dental Photographs: Validation Using Reference Retainers. Journal of Clinical Periodontology. https://doi.org/10.1111/jcpe.14164

Vancouver

Chen YC, Chen L, Lai YL, Chang WT, Lee SY. AI ‐Driven Detection and Measurement of Keratinized Gingiva in Dental Photographs: Validation Using Reference Retainers. Journal of Clinical Periodontology. 2025. doi:10.1111/jcpe.14164

BibTeX
@article{Chen2025,
  author = {Chen, Ya-Chi and Chen, Ling and Lai, Yu-Lin and Chang, Wei-Ting and Lee, Shyh-Yuan},
  title = {AI ‐Driven Detection and Measurement of Keratinized Gingiva in Dental Photographs: Validation Using Reference Retainers},
  journal = {Journal of Clinical Periodontology},
  year = {2025},
  doi = {10.1111/jcpe.14164}
}
Kostenlos starten →