A Comparative Benchmark of Modern Deep Learning Backbones for Multi-Class Periodontal Bone Loss Staging
Journal of Clinical Medicine · 06.11.2026 · 15 (21) · 8644
DOI: 10.3390/jcm15218644 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Clinical Medicine, online erschienen am 06.11.2026. Verfasst von Nurullah Duger und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Periodontology, Faculty of Dentistry, Firat University, Elazig 23100, Turkey. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Fırat University, the Scientific Research Projects Coordination Unit of Firat University.
Abstract (Originalsprache)
Background/Objectives: Early diagnosis of periodontal disease is critical to prevent tissue destruction and tooth loss. This study evaluates six modern deep learning architectures for four-class radiographic periodontal bone loss grading on panoramic radiographs.
Methods
This retrospective study analyzed 1053 images. Two periodontologists independently classified each image according to the most severely affected tooth into four bone loss categories: Healthy, Mild (50%). Six convolutional, transformer, and hybrid backbones (ConvNeXt-Tiny, MaxViT-Tiny, EfficientViT-B1, FocalNet-Tiny, CAFormer-S18, and DaViT-Tiny) were trained on a 70/10/20 stratified split using an NVIDIA Tesla T4 GPU, the AdamW optimizer, Cosine Annealing, and early stopping. Performance was assessed using standard classification metrics, Matthews Correlation Coefficient, Cohen’s and Quadratic Weighted Kappa, Expected Calibration Error, Area Under the Curve (AUC), McNemar’s test, and bootstrap 95% confidence intervals.
Results
FocalNet-Tiny achieved the highest numerical scores among the tested models, reaching 83.41% accuracy, 84.70% balanced accuracy, 83.50% macro-F1, and 0.9610 OvR AUC. The p-values obtained from the comparisons of the models were all above 0.05, indicating that the classification performances of the evaluated models are quite similar. Bootstrap analysis confirmed stable prediction metrics for the top-scoring model.
Conclusions
This study provides a controlled benchmark of modern deep learning backbones for four-class radiographic periodontal bone loss grading. FocalNet-Tiny achieved the highest numerical metrics and stable bootstrap estimates, yet performed comparably to other contemporary architectures such as ConvNeXt-Tiny. Direct radiographic classification shows promise for automated radiographic pre-screening; however, multi-center external validation and integration with clinical periodontal parameters are required before clinical implementation.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Periodontology, Faculty of Dentistry, Firat University, Elazig 23100, Turkey.
- Digital Transformation and Software Office, Firat University, Elazig 23100, Turkey.
- Department of Software Engineering, Malatya Turgut Ozal University, Malatya 44200, Turkey.
- Department of Artificial Intelligence and Data Engineering, Firat University, Elazig 23119, Turkey.
- Department of Periodontics, College of Dentistry, University of Illinois Chicago, Chicago, IL 60607, USA.
ORCID-Profile: Nurullah Duger, Burak Dagtekin, Furkan Talo, Mucahit Karaduman, Muhammed Yildirim
Förderung
- Fırat University
- the Scientific Research Projects Coordination Unit of Firat University
Zitieren
APA
Duger, N., Dagtekin, B., Talo, F., Aydin, M. A., Karaduman, M., Yildirim, M., Yildirim, T. T., & Tozum, T. F. (2026). A Comparative Benchmark of Modern Deep Learning Backbones for Multi-Class Periodontal Bone Loss Staging. Journal of Clinical Medicine, 15(21), 8644. https://doi.org/10.3390/jcm15218644
Vancouver
Duger N, Dagtekin B, Talo F, Aydin MA, Karaduman M, Yildirim M, et al. A Comparative Benchmark of Modern Deep Learning Backbones for Multi-Class Periodontal Bone Loss Staging. Journal of Clinical Medicine. 2026;15(21):8644. doi:10.3390/jcm15218644
BibTeX
@article{Duger2026,
author = {Duger, Nurullah and Dagtekin, Burak and Talo, Furkan and Aydin, Melek Atille and Karaduman, Mucahit and Yildirim, Muhammed and Yildirim, Tuba Talo and Tozum, Tolga Fikret},
title = {A Comparative Benchmark of Modern Deep Learning Backbones for Multi-Class Periodontal Bone Loss Staging},
journal = {Journal of Clinical Medicine},
year = {2026},
volume = {15},
number = {21},
pages = {8644},
doi = {10.3390/jcm15218644}
}
EUVIMED