A Comparative Benchmark of Modern Deep Learning Backbones for Multi-Class Periodontal Bone Loss Staging
5 affiliations
- 1Department of Periodontology, Faculty of Dentistry, Firat University, Elazig 23100, Turkey
- 2Digital Transformation and Software Office, Firat University, Elazig 23100, Turkey
- 3Department of Software Engineering, Malatya Turgut Ozal University, Malatya 44200, Turkey
- 4Department of Artificial Intelligence and Data Engineering, Firat University, Elazig 23119, Turkey
- 5Department of Periodontics, College of Dentistry, University of Illinois Chicago, Chicago, IL 60607, USA
Abstract
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 (<25%), Moderate (25–50%), and Severe (>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.
Related
- Problem Formulation Outweighs Loss-Function Choice in Deep-Learning Cervical Vertebral Maturation Staging: A Multi-Seed Study on an Imbalanced Public Benchmark.
- Comparative Evaluation of Large Language Models in Responding to Common Misconceptions Regarding Periodontal Disease and Oral Hygiene
- Learning Curve and Detection Accuracy of Sentinel Lymph Node Biopsy for Endometrial Cancer Staging: Experience from an ESGO & GESEA Certified Center
- Comparative Evaluation of Large Language Model Interfaces in Third Molar Surgery Complication Scenarios: Response Quality, Clinical Content, Potential Clinical Risk, Readability, and Externally Observable Response Latency—A Cross-Sectional Comparative Benchmark Study
- Comparative diagnostic accuracy of PSMA-PET versus conventional imaging for Initial T/N/M staging of intermediate- to high-risk prostate cancer: a patient-level paired bivariate meta-analysis.
BibTeX
@article{duger2026comparative,
title = {{A Comparative Benchmark of Modern Deep Learning Backbones for Multi-Class Periodontal Bone Loss Staging}},
author = {Nurullah Duger and Burak Dagtekin and Furkan Talo and Melek Atille Aydin and Mucahit Karaduman and Muhammed Yildirim and Tuba Talo Yildirim and Tolga Fikret Tozum},
journal = {Journal of Clinical Medicine},
year = {2026},
month = nov,
doi = {10.3390/jcm15218644},
url = {https://www.mdpi.com/2077-0383/15/21/8644},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - A Comparative Benchmark of Modern Deep Learning Backbones for Multi-Class Periodontal Bone Loss Staging
AU - Nurullah Duger
AU - Burak Dagtekin
AU - Furkan Talo
AU - Melek Atille Aydin
AU - Mucahit Karaduman
AU - Muhammed Yildirim
AU - Tuba Talo Yildirim
AU - Tolga Fikret Tozum
PY - 2026
DA - 2026/11/06
JO - Journal of Clinical Medicine
DO - 10.3390/jcm15218644
UR - https://www.mdpi.com/2077-0383/15/21/8644
AB - 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 (<25%), Moderate (25–50%), and Severe (>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.
SN - 2077-0383
ER -