Görüntü Filtreleme ve Transfer Öğrenme Yaklaşımları Kullanılarak X-Ray Görüntülerinden Düztaban Deformitesinin Tespiti

Öz

Düztabanlık (pes planus), ayağın çökmesi veya ayaktaki kemik ve dokuları belirli bir düzen içinde tutan bağlar ve kaslar gibi yapıların zayıflaması sonucu eğri yapının düzleşmesi ve çeşitli nedenlerle eğri olması olarak tanımlanan bir durumdur. Bu durum tedavi edilmediği takdirde ayakta bozulmaya bağlı baldır, diz, kalça ve bel ağrılarına ve hatta duruş bozukluklarına yol açabilir. Bu çalışmada, X-ışını görüntülerinden düztabanlık tespiti için Dilatasyon filtresini kullanan transfer öğrenmeye dayalı bir yöntem sunulmuştur. X-ışını görüntü veri seti 402 adet düztabanlık görüntüsü ve 440 adet kontrol görüntüsü içermektedir. Görüntü ön işleme için dilatasyon filtrelemesi kullanılmış ve görüntüler dilatasyon yöntemi ile zenginleştirilmiştir. Görüntü ön işleme sonrasında transfer öğrenme yaklaşımları olan DarkNet19, GoogLeNet, DenseNet-201, ResNet-101 ve MobileNetV2 mimarilerinin performansları karşılaştırılmıştır. Performans ölçümleri için holdout yöntemi kullanılmıştır. Deneysel sonuçlar, DenseNet-201 modelinin 0,9802 genel doğruluk ve 0,96 Cohen's Kappa değeri ile en iyi performansı gösterdiğini göstermektedir. Sonuçlar, genişleme filtreleme ve transfer öğrenme yöntemlerinin birleşiminin otomatik düztabanlık tespiti için etkili bir yaklaşım sağladığını göstermektedir. Literatürdeki benzer çalışmalarla karşılaştırıldığında, önerilen modelin doğruluğu önemli ölçüde daha yüksektir.

Anahtar Kelimeler

• Derin öğrenme
• Transfer öğrenme
• Görüntü işleme
• Düztaban
• Görüntü filtreleme

Detection of Flatfoot Deformity from X-Ray Images Using Image Filtering and Transfer Learning Approaches

Abstract

Flatfoot (pes planus) is a condition defined as the flattening of the curved structure as a result of the collapse of the foot or the weakening of the structures, such as ligaments and muscles that hold the bones and tissues in the foot in a certain order and a curve due to various reasons. If left untreated, this condition can lead to calf, knee, hip, and lower back pain and even postural disorders due to foot deterioration. In this study, a transfer learning-based method is presented using the Dilation filter for flatfoot detection from X-ray images. The X-ray image dataset contains 402 flatfoot images and 440 control images. For image preprocessing, dilation filtering is used, and the images are enhanced with the dilation method. After image preprocessing, the performance of transfer learning approaches, DarkNet19, GoogLeNet, DenseNet-201, ResNet-101, and MobileNetV2 architectures, were compared. The holdout method was used for performance measurements. The experimental results show that the DenseNet-201 model performs the best with an overall accuracy of 0.9802 and a Cohen's Kappa value of 0.96. The results show that the combination of dilation filtering and transfer learning methods provides an effective approach for automatic flatfoot detection. Compared to similar studies in the literature, the accuracy of the proposed model is significantly higher.

Keywords

• Deep learning
• Transfer learning
• Image processing
• Flatfoot
• Image filter

References

1. [1] T. Çit, D. Yılmaz, T. Çaviş, E. Alp, and H. Aydın, “A Software Tool Developed for Automatic Diagnosis of Pes Planus,” in 11th International May 19 Innovative Scientific Approaches Congress, Samsun, Turkey, 2024, pp. 303–311.
2. [2] A. Azizov and Ö. Şevgin, “Pediatrik Pes Planus ve Fizyoterapi.”
3. [3] N. Katz, “The impact of pain management on quality of life,” J. Pain Symptom Manage., vol. 24, no. 1 Suppl, pp. S38–S47, Jan. 2002.
4. [4] S. Erkuş and Ö. Kalenderer, “Pes planovalgus,” Totbid Dergisi, vol. 16, pp. 413–425, 2017.
5. [5] C. J. Lin, K. A. Lai, T. S. Kuan, and Y. L. Chou, “Correlating factors and clinical significance of flexible flatfoot in preschool children,” J. Pediatr. Orthop., vol. 21, no. 3, pp. 378–382, 2001.
6. [6] D. Bordin, G. De Giorgi, G. Mazzocco, and F. Rigon, “Flat and cavus foot, indexes of obesity and overweight in a population of primary-school children,” Minerva Pediatr., vol. 53, no. 1, pp. 7–13, 2001.
7. [7] C. Wang et al., “An efficient local binary pattern based plantar pressure optical sensor image classification using convolutional neural networks,” Optik, vol. 185, pp. 543–557, 2019.
8. [8] Y. Kim and N. Kim, “Deep learning-based pes planus classification model using transfer learning,” Journal of The Korea Society of Computer and Information, vol. 26, no. 4, pp. 21–28, 2021.

9. [9] J. Y. Jung, C. M. Yang, and J. J. Kim, “Decision Tree-Based Foot Orthosis Prescription for Patients with Pes Planus,” International Journal of Environmental Research and Public Health, vol. 19, no. 19, p. 12484, 2022. [Online]. Available: http://www.journalwebsite.com DOI: 10.3390/ijerph191912484.
10. [10] J. Lauder et al., “A fully automatic system to assess foot collapse on lateral weight-bearing foot radiographs: A pilot study,” Computer Methods and Programs in Biomedicine, vol. 213, p. 106507, 2022. DOI: 10.1016/j.cmpb.2022.106507.
11. [11] S. M. Ryu et al., “Automated landmark identification for diagnosis of the deformity using a cascade convolutional neural network (FlatNet) on weight-bearing lateral radiographs of the foot,” Computers in Biology and Medicine, vol. 148, p. 105914, 2022. DOI: 10.1016/j.compbiomed.2022.105914.
12. [12] S. M. Ryu, K. Shin, S. W. Shin, S. Lee and N. Kim, “Enhancement of evaluating flatfoot on a weight-bearing lateral radiograph of the foot with U-Net based semantic segmentation on the long axis of tarsal and metatarsal bones in an active learning manner,” Computers in Biology and Medicine, vol. 145, p. 105400, 2022. DOI: 10.1016/j.compbiomed.2022.105400.
13. [13] Y. Gül, S. Yaman, D. Avcı, A. H. Çilengir, M. Balaban, and H. Güler, “A novel deep transfer learning-based approach for automated Pes Planus diagnosis using X-ray image,” Diagnostics, vol. 13, no. 9, p. 1662, 2023. DOI: 10.3390/diagnostics13091662. Dataset from: https://www.kaggle.com/datasets/suleyman32/pesplanus-two-class-dataset
14. [14] F. A. Alsaidi and K. M. Moria, “Flatfeet Severity-Level Detection Based on Alignment Measuring,” Sensors, vol. 23, no. 19, p. 8219, 2023. DOI: 10.3390/s23198219.
15. [15] K. Doğan, T. Selçuk, and A. Yılmaz, “A Novel Model Based on CNN–ViT Fusion and Ensemble Learning for the Automatic Detection of Pes Planus,” Journal of Clinical Medicine, vol. 13, no. 16, p. 4800, 2024. DOI: 10.3390/jcm13164800.
16. [16] J. Redmon and A. Farhadi, “YOLO9000: better, faster, stronger,” in IEEE Conference on Computer Vision and Pattern Recognition, 2017, pp. 7263–7271.
17. [17] M. Lin, “Network in network,” arXiv preprint arXiv:1312.4400, 2013. [Online]. Available: http://arxiv.org/abs/1312.4400.
18. [18] Z. Zhou, Y. Hu, Z. Zhu, and Y. Wang, “Fabric wrinkle objective evaluation model with random vector function link based on optimized artificial hummingbird algorithm,” Journal of Natural Fibers, vol. 20, no. 1, p. 2163026, 2023. DOI: 10.1080/15440478.2023.2163026.
19. [19] G. Huang, Z. Liu, L. Van Der Maaten, and K. Q. Weinberger, “Densely connected convolutional networks,” in IEEE Conference on Computer Vision and Pattern Recognition, 2017, pp. 4700–4708.
20. [20] V. Miglani and M. P. S. Bhatia, “Skin lesion classification: A transfer learning approach using efficientnets,” in International Conference on Advanced Machine Learning Technologies and Applications, Singapore, 2020, pp. 315–324.
21. [21] A. Lumini and L. Nanni, “Deep learning and transfer learning features for plankton classification,” Ecological Informatics, vol. 51, pp. 33–43, 2019. DOI: 10.1016/j.ecoinf.2019.02.007.
22. [22] H. Kumar, A. Virmani, S. Tripathi, R. Agrawal and S. Kumar, “Transfer learning and supervised machine learning approach for detection of skin cancer: performance analysis and comparison,” Transfer, vol. 10, no. 1, 2021. DOI: 10.1007/s11203-021-00207-3.
23. [23] M. B. Özküçük, Ö. F. Alçin and M. T. Gençoğlu, (2024). Transfer öğrenme yaklaşımı kullanılarak izolatör kusurlarının tespiti. Dicle Üniversitesi Mühendislik Fakültesi Mühendislik Dergisi, 15(2), 323-330. DOI: 10.24012/dumf.1415322.
24. [24] Z. Guo, Q. Chen, G. Wu, Y. Xu, R. Shibasaki, and X. Shao, “Village building identification based on ensemble convolutional neural networks,” Sensors, vol. 17, no. 11, p. 2487, 2017. DOI: 10.3390/s17112487.
25. [25] Y. Zou, L. Wu, C. Zuo, L. Chen, B. Zhou, and H. Zhang, “White blood cell classification network using MobileNetv2 with multiscale feature extraction module and attention mechanism,” Biomedical Signal Processing and Control, vol. 99, p. 106820, 2025. DOI: 10.1016/j.bspc.2024.106820.
26. [26] F. Arnia, K. Saddami, and K. Munadi, “Dcnet: Noise-robust convolutional neural networks for degradation classification on ancient documents,” Journal of Imaging, vol. 7, no. 7, p. 114, 2021. DOI: 10.3390/jimaging7070114.
27. [27] A. Ciran, and E. Özbay, “Derin Öğrenme ve Özellik Seçimi Yaklaşımları Kullanılarak Göz Hastalıkları Tespiti,” Dicle Üniversitesi Mühendislik Fakültesi Mühendislik Dergisi, vol. 15, no. 2, pp. 421-433, 2024. DOI: 10.24012/dumf.1465929
28. [28] S. Targ, D. Almeida, and K. Lyman, “Resnet in resnet: Generalizing residual architectures,” arXiv preprint arXiv:1603.08029, 2016. [Online]. Available: http://arxiv.org/abs/1603.08029.
29. [29] S. Keskin, O. Sevli, and E. Okatan, “Comparative analysis of the classification of recyclable wastes,” Journal of Scientific Reports-A, vol. (055), pp. 70-79, 2023. DOI: 10.59313/jsr-a.1335276
30. [30] S. Albahli, and T. Nazir, “AI-CenterNet CXR: An artificial intelligence (AI) enabled system for localization and classification of chest X-ray disease,” Frontiers in Medicine, vol. 9, p. 955765, 2022. DOI: 10.3389/fmed.2022.955765.
31. [31] A. Alan and M. Karabatak, “Veri seti-sınıflandırma ilişkisinde performansa etki eden faktörlerin değerlendirilmesi,” Fırat Üniversitesi Mühendislik Bilimleri Dergisi, vol. 32, no. 2, pp. 531–540, 2020.
32. [32] C. Hark, “Sahte haber tespiti için derin bağlamsal kelime gömülmeleri ve sinirsel ağların performans değerlendirmesi,” Fırat Üniversitesi Mühendislik Bilimleri Dergisi, vol. 34, no. 2, pp. 733–742, 2022. DOI: 10.35234/fumbd.1126688
33. [33] S. Çelik and Ö. Kasım, “Detection of tumor slice in brain magnetic resonance images by feature optimized transfer learning,” Aksaray University Journal of Science and Engineering, vol. 4, no. 2, pp. 187–198, 2020. DOI: 10.29002/asujse.820599.
34. [34] S. Akyol, M. Yıldırım, and B. Alataş, “Multi-feature fusion and improved BO and IGWO metaheuristics-based models for automatically diagnosing the sleep disorders from sleep sounds,” Computers in Biology and Medicine, vol. 157, p. 106768, 2023. DOI: 10.1016/j.compbiomed.2023.106768.
35. [35] Ö. B. Bilgen and N. Doğan, “Puanlayıcılar arası güvenirlik belirleme tekniklerinin karşılaştırılması,” Journal of Measurement and Evaluation in Education and Psychology, vol. 8, no. 1, pp. 63–78, 2017. DOI: 10.21031/epod.294847.
36. [36] N. W. S. Wardhani, M. Y. Rochayani, A. Iriany, A. D. Sulistyono, and P. Lestantyo, “Cross-validation metrics for evaluating classification performance on imbalanced data,” 2019 International Conference on Computer, Control, Informatics, 2019.