Conference Paper2025Cited by 2

Comparative evaluation of deep learning models for multi-disease classification using reflective dental imaging

Mohammed Tashfiq Islam, Mustaqueem Alam, JM Sadik-Ul Islam Smaron, Jahanggir Hossain Setu, Siam Tahsin Bhuiyan, Rashedur Rahman, Asif Mahmud, Saadia Binte Alam, Ashraful Islam

2025 IEEE 23rd Student Conference on Research and Development (SCOReD)

IEEE, pp. 1-6

  • Good Health and Well-being — Relevance: 94%
DOI: 10.1109/scored68498.2025.11399140URL

Abstract

Automated and accurate classification of common dental conditions from ordinary Red-Green-Blue (RGB) images can enable low-cost, scalable screening and teledentistry in resource-limited settings. This study evaluates and compares five distinct Convolutional Neural Network (CNN) architectures, including Residual Network (ResNet)-50, ResNet-101, Dense Convolutional Network (DenseNet)-121, MobileNetV2, and Neural Architecture Search Network (NASNet)-Mobile, for the multiclass classification of reflective dental images. A Kaggle-sourced dataset of 5,048 images, labeled in five categories (Dental Calculus, Dental Caries, Hypodontia, Mouth Ulcer, and Tooth Discoloration), was processed using a preprocessing pipeline that included Contrast Limited Adaptive Histogram Equalization (CLAHE) for contrast enhancement and class-aware augmentation. Performance was assessed with accuracy, precision, recall, and F1-score to address class imbalance. ResNet-101 delivered the best overall performance after proper fine-tuning (accuracy ≈ 97%; macro F1-Score ≈ 94%), followed closely by the other models. Additionally, dental caries and small / low-contrast lesions were identified as common failure modes by means of qualitative analysis through visual inspection and saliency maps. Overall, the results suggest that using deep architectures to capture dental features from reflective images is feasible, indicating a pathway for low-cost screening and teledentistry. Future work will increase the heterogeneity of the dataset, increase sensitivity to low-contrast lesions, and study model compression to enable local deployment on the device.

Keywords

  • dental caries
  • tooth discoloration
  • MobileNetV2
  • tooth agenesis
  • DenseNet121
  • ResNet-101
  • class imbalance
  • teledentistry
  • dental calculus
  • saliency maps

CCDS Authors

Siam Tahsin Bhuiyan
Siam Tahsin Bhuiyan

Siam Tahsin Bhuiyan

  • Intern, CCDS

Bachelors

Independent University, Bangladesh

Published at

Scientific Reports

Md. Rashedur Rahman, D. Eng.
Md. Rashedur Rahman, D. Eng.

Md. Rashedur Rahman, D. Eng.

  • Co-Director, MIRA & NEST Wings

Assistant Professor

Department of Computer Science and Engineering

Independent University, Bangladesh

Published at

CHI, HCII, Scientific Reports, Cognitive Computation, Neural Computing and Applications

Asif Mahmud
Asif Mahmud

Asif Mahmud

  • Co-Director, IPT Wing

Lecturer

Department of Computer Science and Engineering

Independent University, Bangladesh

Saadia Binte Alam, PhD
Saadia Binte Alam, PhD

Saadia Binte Alam, PhD

  • Director, MIRA & WiSE Wings

Associate Professor

Department of Computer Science and Engineering

Independent University, Bangladesh

Published at

HCII, Scientific Reports, Cognitive Computation, Neural Computing and Applications

Ashraful Islam, PhD
Ashraful Islam

Ashraful Islam, PhD

  • Center Director, CCDS
  • Director, HCI Wing

Assistant Professor

Department of Computer Science and Engineering

Independent University, Bangladesh (IUB)

Published at

CHI, UIST, UbiComp, IJCNN, ICIP, HCII, ICDMW, BLP, IEEE Internet of Things Journal, JMIR, JMIR Human Factors, JMIR Formative Research, IEEE Access, Smart Agricultural Technology, Diabetes Research and Clinical Practice, PLOS ONE

Research Interests

Human-Computer Interaction, AI for Social Good, AI for Public Health, AI for Impact

References

  1. 1.Joachim Krois, Thomas Ekert, Leonie Meinhold, Tatiana Golla, Basel Kharbot, Agnes Wittemeier, Christof Dörfer, Falk Schwendicke. (2019). Deep Learning for the Radiographic Detection of Periodontal Bone Loss. Scientific Reports, 9(1), 8495[10.1038/s41598-019-44839-3]
  2. 2.Yassir Edrees Almalki, Amsa Imam Din, Muhammad Ramzan, Muhammad Irfan, Khalid Mahmood Aamir, Abdullah Almalki, Saud Alotaibi, Ghada Alaglan, Hassan A Alshamrani, Saifur Rahman. (2022). Deep Learning Models for Classification of Dental Diseases Using Orthopantomography X-ray OPG Images. Sensors, 22(19), 7370[10.3390/s22197370]
  3. 3.Ivane Delos Santos Chen, Chieh-Ming Yang, Mei-Juan Chen, Mei-Juan Chen, Ming-Chin Chen, Ming-Chin Chen, Ro-Min Weng, Chia-Hung Yeh. (2023). Deep Learning-Based Recognition of Periodontitis and Dental Caries in Dental X-ray Images. Bioengineering, 10(8), 911[10.3390/bioengineering10080911]
  4. 4.Yasmin M. Alsakar, Naira Elazab, Nermeen Nader, Waleed Mohamed, Mohamed Ezzat, Mohammed Elmogy. (2024). Multi-label dental disorder diagnosis based on MobileNetV2 and swin transformer using bagging ensemble classifier. Scientific Reports, 14(1), 25193[10.1038/s41598-024-73297-9]
  5. 5.Saron Park, Habibilloh Erkinov, Md. Al Mehedi Hasan, Seoul-Hee Nam, Yu-Rin Kim, Jungpil Shin, Won-Du Chang. (2023). Periodontal Disease Classification with Color Teeth Images Using Convolutional Neural Networks. Electronics, 12(7), 1518[10.3390/electronics12071518]
  6. 6.Alfonso Ramírez-Pedraza, Sebastián Salazar-Colores, Crystel Cardenas-Valle, Juan Terven, José-Joel González-Barbosa, Francisco-Javier Ornelas-Rodriguez, Juan-Bautista Hurtado-Ramos, Raymundo Ramirez-Pedraza, Diana-Margarita Córdova-Esparza, Julio-Alejandro Romero-González. (2025). Deep Learning in Oral Hygiene: Automated Dental Plaque Detection via YOLO Frameworks and Quantification Using the O’Leary Index. Diagnostics, 15(2), 231[10.3390/diagnostics15020231]
  7. 7.Viduni Liyanage, Mengqiu Tao, Joon Soo Park, Kate N Wang, Somayyeh Azimi. (2023). Malignant and non-malignant oral lesions classification and diagnosis with deep neural networks. Journal of Dentistry, 137, 104657[10.1016/j.jdent.2023.104657]
  8. 8.Xiao Wei, Liu Chanjuan, Jiang Ke, Ye Linyun, Gao Jinxing, Wang Quanbing. (2024). Convolutional neural network for oral cancer detection combined with improved tunicate swarm algorithm to detect oral cancer. Scientific Reports, 14(1), 28675[10.1038/s41598-024-79250-0]
  9. 9.Zuhal Can, Sahin Isik, Yildiray Anagun. (2024). CVApool: using null-space of CNN weights for the tooth disease classification. Neural Computing and Applications, 36(26), 16567–16579[10.1007/s00521-024-09995-2]
  10. 10.Yilin Li, Jun Zhang. (2024). Multi-Label Dental Image Classification via Vision Transformer for Orthopantomography X-ray Images. Computer-Aided Design and Applications, 198–207[10.14733/cadaps.2024.s21.198-207]
  11. 11.Yujie Ma, Maged Ali Al-Aroomi, Yutian Zheng, Wenjie Ren, Peixuan Liu, Qing Wu, Ye Liang, Canhua Jiang. (2025). Application of Mask R-CNN for automatic recognition of teeth and caries in cone-beam computerized tomography. BMC Oral Health, 25(1), 927[10.1186/s12903-025-06293-8]
  12. 12.Manoj Jaiswal, Megha Sharma, Padmavati Khandnor, Ashima Goyal, Rajendra Belokar, Sandeep Harit, Tamanna Sood, Kannav Goyal, Pallavi Dua. (2023). Deep Learning Models for Classification of Deciduous and Permanent Teeth From Digital Panoramic Images. Cureus, 15(12), e49937[10.7759/cureus.49937]

Cited By

  1. 1.Mustaqueem Alam, J. M. Sadik-Ul Islam Smaron, Rashedur Rahman, Md Asif Bin Khaled, Ashraful Islam, M. Ashraful Amin. (2026). Evaluating the Impact of Image Preprocessing on Deep Learning-Based Fetal Ultrasound Classification Under Class Imbalance. , 128–135[10.1109/icicds70526.2026.11604754]
  2. 2.Sayeda Subha Ashrafi, Tushar Ahmed Shishir. (2026). Deep Learning-Based Automated Breast Cancer Detection from Histopathology Images: Benchmarking CNN and Transformer Architectures. , 1–6[10.1109/peccii70991.2026.11662018]
← All publications