Transformer-Based Otitis Media Classification: A Comparative Study of ViT, DeiT, and PVT Architectures in Otoscopic Image Analysis

Sedat Örenç

doi:10.17678/beuscitech.1920012

Transformer-Based Otitis Media Classification: A Comparative Study of ViT, DeiT, and PVT Architectures in Otoscopic Image Analysis

Abstract

Otitis media (OM) is one of the most prevalent middle ear diseases globally. It represents a major clinical and public health burden. It imposes significant health and socio-economic challenges, particularly among pediatric populations. Conventional diagnostic tools, such as otoscopy and tympanometry, are constrained by inherent subjectivity and operator dependency. Furthermore, they suffer from low specificity. This underscores the critical need for reliable, scalable, and automated diagnostic solutions. Recent advances in deep learning have improved image-based diagnosis; however, traditional Convolutional Neural Networks (CNNs) remain constrained by their inability to capture long-range dependencies essential for distinguishing subtle tympanic membrane pathologies. To address these limitations, this study systematically benchmarks three transformer-based architectures: Vision Transformer (ViT), Data-Efficient Image Transformer (DeiT), and Pyramid Vision Transformer (PvT). The objective is to automate the classification of otoscopic images. Classification targets include Acute Otitis Media (AOM), Chronic Otitis Media (COM), and normal cases. A balanced dataset of 1,800 images was curated and augmented. This ensured a fair evaluation under standardized training conditions. The experimental results demonstrate that ViT performs well for AOM, achieving an accuracy and F1-score of 0.98. However, its performance declines for chronic and normal cases. In contrast, DeiT produces the most consistent results across all categories, achieving near-perfect accuracy of 1.00 for acute OM cases and 0.96 for chronic and normal cases. PvT also demonstrates strong performance, achieving an accuracy of 1.00 for OM and 0.99 for normal. These findings demonstrate the superior robustness and clinical potential of DeiT and PvT compared to ViT, suggesting their suitability for real-world applications. Beyond delivering a reproducible benchmarking framework, this work contributes toward bridging the gap between algorithmic innovation and clinical translation. This provides a way of achieving reliable, interpretable and scalable AI-assisted diagnosis of otitis media.

Keywords

Ethical Statement

The author declare that this study complies with research and publication ethics. This study does not require an ethics committee approval as it does not involve any human or animal subjects, surveys, or clinical data.

References

I. A. Mohamed, Z. A. Mohamed, F. Ning, and W. Xin, “The Prevalence and Risk Factors Associated with Otitis Media in Children under Five Years of Age in Mogadishu, Somalia: A Hospital-Based Cross-Sectional Study,” International Journal of Otolaryngology and Head & Neck Surgery, vol. 12, no. 06, pp. 426–443, 2023, doi: 10.4236/ijohns.2023.126046.
M. W. You et al., “The Roles of NOD-like Receptors in Innate Immunity in Otitis Media,” Feb. 01, 2022, MDPI. doi: 10.3390/ijms23042350.
L. Cupples, T. Y. C. Ching, and S. Hou, “Speech, language, functional communication, psychosocial outcomes and QOL in school-age children with congenital unilateral hearing loss,” Front Pediatr, vol. 12, Mar. 2024, doi: 10.3389/fped.2024.1282952.
F. Folino, M. Caruso, P. Bosi, M. Aldè, S. Torretta, and P. Marchisio, “Acute otitis media diagnosis in childhood: still a problem in 2023?,” Ital J Pediatr, vol. 50, no. 1, Dec. 2024, doi: 10.1186/s13052-024-01588-y.
J. H. Stephens et al., “A multimodal machine learning algorithm improved diagnostic accuracy for otitis media in a school aged Aboriginal population,” J Biomed Inform, vol. 164, Apr. 2025, doi: 10.1016/j.jbi.2025.104801.
R. G. Kashani et al., “Shortwave infrared otoscopy for diagnosis of middle ear effusions: a machine-learning-based approach,” Sci Rep, vol. 11, no. 1, Dec. 2021, doi: 10.1038/s41598-021-91736-9.
C. B. Botelho, R. Tortoriello, S. N. Koch, and J. I. Fernandes, “Clinical aspects of middle ear tympanokeratoma in dogs diagnosed through advanced imaging, otoendoscopy and histopathological evaluation,” Vet Dermatol, vol. 36, no. 4, pp. 417–423, Aug. 2025, doi: 10.1111/vde.13321.
D. Muhammad and M. Bendechache, “Unveiling the black box: A systematic review of Explainable Artificial Intelligence in medical image analysis,” Dec. 01, 2024, Elsevier B.V. doi: 10.1016/j.csbj.2024.08.005.

S. Örenç, E. Acar, M. S. Özerdem, S. Şahin, and A. Kaya, “Automatic Identification of Adenoid Hypertrophy via Ensemble Deep Learning Models Employing X-ray Adenoid Images,” Journal of Imaging Informatics in Medicine, 2025, doi: 10.1007/s10278-025-01423-8.
S. Orenc, M. S. Ozerdem, E. Acar, and M. Yilmaz, “Automatic segmentation of chest X-ray images via deep-improved various U-Net techniques,” Digit Health, vol. 11, May 2025, doi: 10.1177/20552076251366855.
Y. Choi, J. Chae, K. Park, J. Hur, J. Kweon, and J. H. Ahn, “Automated multi-class classification for prediction of tympanic membrane changes with deep learning models,” PLoS One, vol. 17, no. 10 October, Oct. 2022, doi: 10.1371/journal.pone.0275846.
A. Younesi, M. Ansari, M. Fazli, A. Ejlali, M. Shafique, and J. Henkel, “A Comprehensive Survey of Convolutions in Deep Learning: Applications, Challenges, and Future Trends,” Feb. 2024, [Online]. Available: http://arxiv.org/abs/2402.15490
Y. Lu, S. Wang, B. Wang, X. Zhang, X. Wang, and Y. Zhao, “Enhanced Window-Based Self-Attention with Global and Multi-Scale Representations for Remote Sensing Image Super-Resolution,” Remote Sens (Basel), vol. 16, no. 15, Aug. 2024, doi: 10.3390/rs16152837.
A. W. Salehi et al., “A Study of CNN and Transfer Learning in Medical Imaging: Advantages, Challenges, Future Scope,” Apr. 01, 2023, MDPI. doi: 10.3390/su15075930.
A. Khan et al., “A survey of the vision transformers and their CNN-transformer based variants,” Artif Intell Rev, vol. 56, pp. 2917–2970, Dec. 2023, doi: 10.1007/s10462-023-10595-0.
Y. Wang, Y. Deng, Y. Zheng, P. Chattopadhyay, and L. Wang, “Vision Transformers for Image Classification: A Comparative Survey,” Jan. 01, 2025, Multidisciplinary Digital Publishing Institute (MDPI). doi: 10.3390/technologies13010032.
Z. Cömert, A. Sbrollini, F. Demircan, and L. Burattini, “Computerized otoscopy image-based artificial intelligence model utilizing deep features provided by vision transformer, grid search optimization, and support vector machine for otitis media diagnosis,” Neural Comput Appl, vol. 36, no. 36, pp. 23113–23129, Dec. 2024, doi: 10.1007/s00521-024-10457-y.
F. Demircan, M. Ekinci, and Z. Cömert, “Enhancing intra-aural disease classification with attention-based deep learning models,” Neural Comput Appl, vol. 37, no. 9, pp. 6601–6616, Mar. 2025, doi: 10.1007/s00521-025-10990-4.
B. Ananda Deva, V. Sood, A. Fahim, P. M. Arjun Krishna, and T. Anjali, “Feature Fusion: Enhancing Middle Ear Infection Diagnosis with Vision Transformer and Symptom Encoding,” in Proceedings of 8th International Conference on Inventive Computation Technologies, ICICT 2025, Institute of Electrical and Electronics Engineers Inc., 2025, pp. 233–239. doi: 10.1109/ICICT64420.2025.11005139.
E. A. Khan and A. A. H. Khan, “Sumotosima: A Framework and Dataset for Classifying and Summarizing Otoscopic Images,” Aug. 2024, [Online]. Available: http://arxiv.org/abs/2408.06755
C. Dubois et al., “Development and validation of a smartphone-based deep-learning-enabled system to detect middle-ear conditions in otoscopic images,” NPJ Digit Med, vol. 7, no. 1, Dec. 2024, doi: 10.1038/s41746-024-01159-9.
A. Dosovitskiy et al., “An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale,” Jun. 2021, [Online]. Available: http://arxiv.org/abs/2010.11929
J. Maurício, I. Domingues, and J. Bernardino, “Comparing Vision Transformers and Convolutional Neural Networks for Image Classification: A Literature Review,” May 01, 2023, MDPI. doi: 10.3390/app13095521.
R. Azad et al., “Advances in Medical Image Analysis with Vision Transformers: A Comprehensive Review,” Nov. 2023, [Online]. Available: http://arxiv.org/abs/2301.03505
B. Guo et al., “An enhanced network for extracting tunnel lining defects using transformer encoder and aggregate decoder,” International Journal of Applied Earth Observation and Geoinformation, vol. 136, Feb. 2025, doi: 10.1016/j.jag.2024.104259.
H. Touvron, M. Cord, and H. Jégou, “DeiT III: Revenge of the ViT.”
M. Asante, G. Adbdul-Salam, B. Nkrumah, and W. K. Adu-Gyamfi, “Data-Efficient Image Transformers for Robust Malware Family Classification,” Journal of Cyber Security, vol. 6, no. 1, pp. 131–153, 2024, doi: 10.32604/jcs.2024.053954.
W. Wang et al., “Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions.” [Online]. Available: https://github.com/whai362/PVT

Details

Primary Language

English

Subjects

Image Processing

Journal Section

Research Article

Authors

Sedat Örenç ^*
0000-0002-1190-2849
Türkiye

Publication Date

June 30, 2026

Submission Date

March 31, 2026

Acceptance Date

June 3, 2026

Published in Issue

Year 2026 Volume: 16 Number: 1

DOI

https://doi.org/10.17678/beuscitech.1920012

IZ

https://izlik.org/JA44WX64LD

Cite

RIS / Bibtex

APA

Örenç, S. (2026). Transformer-Based Otitis Media Classification: A Comparative Study of ViT, DeiT, and PVT Architectures in Otoscopic Image Analysis. Bitlis Eren University Journal of Science and Technology, 16(1), 53-73. https://doi.org/10.17678/beuscitech.1920012

AMA

1.Örenç S. Transformer-Based Otitis Media Classification: A Comparative Study of ViT, DeiT, and PVT Architectures in Otoscopic Image Analysis. Bitlis Eren University Journal of Science and Technology. 2026;16(1):53-73. doi:10.17678/beuscitech.1920012

Chicago

Örenç, Sedat. 2026. “Transformer-Based Otitis Media Classification: A Comparative Study of ViT, DeiT, and PVT Architectures in Otoscopic Image Analysis”. Bitlis Eren University Journal of Science and Technology 16 (1): 53-73. https://doi.org/10.17678/beuscitech.1920012.

EndNote

Örenç S (June 1, 2026) Transformer-Based Otitis Media Classification: A Comparative Study of ViT, DeiT, and PVT Architectures in Otoscopic Image Analysis. Bitlis Eren University Journal of Science and Technology 16 1 53–73.

IEEE

[1]S. Örenç, “Transformer-Based Otitis Media Classification: A Comparative Study of ViT, DeiT, and PVT Architectures in Otoscopic Image Analysis”, Bitlis Eren University Journal of Science and Technology, vol. 16, no. 1, pp. 53–73, June 2026, doi: 10.17678/beuscitech.1920012.

ISNAD

Örenç, Sedat. “Transformer-Based Otitis Media Classification: A Comparative Study of ViT, DeiT, and PVT Architectures in Otoscopic Image Analysis”. Bitlis Eren University Journal of Science and Technology 16/1 (June 1, 2026): 53-73. https://doi.org/10.17678/beuscitech.1920012.

JAMA

1.Örenç S. Transformer-Based Otitis Media Classification: A Comparative Study of ViT, DeiT, and PVT Architectures in Otoscopic Image Analysis. Bitlis Eren University Journal of Science and Technology. 2026;16:53–73.

MLA

Örenç, Sedat. “Transformer-Based Otitis Media Classification: A Comparative Study of ViT, DeiT, and PVT Architectures in Otoscopic Image Analysis”. Bitlis Eren University Journal of Science and Technology, vol. 16, no. 1, June 2026, pp. 53-73, doi:10.17678/beuscitech.1920012.

Vancouver

1.Sedat Örenç. Transformer-Based Otitis Media Classification: A Comparative Study of ViT, DeiT, and PVT Architectures in Otoscopic Image Analysis. Bitlis Eren University Journal of Science and Technology. 2026 Jun. 1;16(1):53-7. doi:10.17678/beuscitech.1920012