Klasifikasi Ulkus Kaki Diabetik Berbasis Transfer Learning: Studi Komparatif Arsitektur CNN

Nur Nafiiyah; Rifqi Firdaus Al Jauhari; Agus Harjoko; Achmad Nizar Hidayanto

doi:10.30865/json.v7i3.9580

Authors

Nur Nafiiyah Universitas Islam Lamongan
Rifqi Firdaus Al Jauhari Universitas Islam Lamongan
Agus Harjoko Universitas Gadjah Mada
Achmad Nizar Hidayanto Universitas Indonesia

DOI:

https://doi.org/10.30865/json.v7i3.9580

Keywords:

Diabetes, DFU, Convolutional Neural Network, Transfer Learning, Ulkus Kaki

Abstract

Proses klasifikasi citra medis secara otomatis sangat penting untuk membantu proses diagnosis yang cepat dan objektif. Deep learning, terutama Convolutional Neural Network (CNN), telah menunjukkan kemampuan yang lebih baik untuk mengidentifikasi dan mengklasifikasikan gambar. Namun, model yang tepat masih sulit dipilih karena perbedaan dalam desain arsitektur CNN dapat berdampak besar pada kinerja. Tujuan dari penelitian ini adalah untuk mengevaluasi dan membandingkan kinerja lima arsitektur CNN (VGG16, VGG19, ResNet101, InceptionV3, dan DenseNet121) ketika menggunakan dataset Ulkus Kaki Diabetes (Diabetic Foot Ulcer/DFU) untuk mengklasifikasikan gambar medis kelas normal dan abnormal. Metode berbasis dataset digunakan untuk menjalankan semua eksperimen. Akurasi, ketepatan, recall, dan F1-score digunakan untuk menilai model. Hasil eksperimen menunjukkan bahwa ResNet101 adalah yang terbaik dengan akurasi dan F1-score sebesar 99,87%, diikuti oleh VGG19 dan VGG16, masing-masing dengan F1-score di atas 99%. DenseNet121 dan InceptionV3 juga menunjukkan kinerja yang kompetitif, meskipun sedikit di bawah model berbasis residual dan VGG. Hasil ini menunjukkan betapa pentingnya melakukan analisis komparatif saat memilih arsitektur CNN untuk klasifikasi citra medis berbasis deep learning.

References

L. Alzubaidi, M. A. Fadhel, S. R. Oleiwi, O. Al-Shamma, and J. Zhang, “DFU_QUTNet: diabetic foot ulcer classification using novel deep convolutional neural network,” Multimed. Tools Appl., vol. 79, no. 21–22, 2020, doi: 10.1007/s11042-019-07820-w.

M. S. A. Toofanee et al., “DFU-SIAM a Novel Diabetic Foot Ulcer Classification With Deep Learning,” IEEE Access, vol. 11, 2023, doi: 10.1109/ACCESS.2023.3312531.

S. Biswas, R. Mostafiz, B. K. Paul, K. M. M. Uddin, M. A. Hadi, and F. Khanom, “DFU_XAI: A Deep Learning-Based Approach to Diabetic Foot Ulcer Detection Using Feature Explainability,” Biomedical Materials and Devices, vol. 2, no. 2, 2024, doi: 10.1007/s44174-024-00165-5.

H. A. Ismael, N. H. Al-A’araji, and B. K. Shukur, “An Enhanced Diabetic Foot Ulcer Classification Approach Using GLCM and Deep Convolution Neural Network,” Karbala International Journal of Modern Science, vol. 8, no. 4, 2022, doi: 10.33640/2405-609X.3268.

N. Al-Garaawi, R. Ebsim, A. F. H. Alharan, and M. H. Yap, “Diabetic foot ulcer classification using mapped binary patterns and convolutional neural networks,” Comput. Biol. Med., vol. 140, 2022, doi: 10.1016/j.compbiomed.2021.105055.

R. Nanda, A. Nath, S. Patel, and E. Mohapatra, “Machine learning algorithm to evaluate risk factors of diabetic foot ulcers and its severity,” Med. Biol. Eng. Comput., vol. 60, no. 8, 2022, doi: 10.1007/s11517-022-02617-w.

M. H. Yap, B. Cassidy, J. M. Pappachan, C. O’Shea, D. Gillespie, and N. D. Reeves, “Analysis towards classification of infection and ischaemia of diabetic foot ulcers,” in BHI 2021 - 2021 IEEE EMBS International Conference on Biomedical and Health Informatics, Proceedings, 2021. doi: 10.1109/BHI50953.2021.9508563.

M. Ahsan, S. Naz, R. Ahmad, H. Ehsan, and A. Sikandar, “A Deep Learning Approach for Diabetic Foot Ulcer Classification and Recognition,” Information (Switzerland), vol. 14, no. 1, 2023, doi: 10.3390/info14010036.

S. Nagaraju et al., “Automated Diabetic Foot Ulcer Detection and Classification Using Deep Learning,” IEEE Access, vol. 11, 2023, doi: 10.1109/ACCESS.2023.3332292.

P. Xie et al., “An explainable machine learning model for predicting in-hospital amputation rate of patients with diabetic foot ulcer,” Int. Wound J., vol. 19, no. 4, 2022, doi: 10.1111/iwj.13691.

A. Khandakar et al., “A Novel Machine Learning Approach for Severity Classification of Diabetic Foot Complications Using Thermogram Images,” Sensors, vol. 22, no. 11, 2022, doi: 10.3390/s22114249.

M. A. Fadhel, L. Alzubaidi, Y. Gu, J. Santamaría, and Y. Duan, “Real-time diabetic foot ulcer classification based on deep learning & parallel hardware computational tools,” Multimed. Tools Appl., vol. 83, no. 27, 2024, doi: 10.1007/s11042-024-18304-x.

Z. Schäfer, A. Mathisen, K. Svendsen, S. Engberg, T. Rolighed Thomsen, and K. Kirketerp-Møller, “Toward Machine-Learning-Based Decision Support in Diabetes Care: A Risk Stratification Study on Diabetic Foot Ulcer and Amputation,” Front. Med. (Lausanne)., vol. 7, 2021, doi: 10.3389/fmed.2020.601602.

M. H. Yap et al., “Deep learning in diabetic foot ulcers detection: A comprehensive evaluation,” Comput. Biol. Med., vol. 135, 2021, doi: 10.1016/j.compbiomed.2021.104596.

S. K. Das, S. Namasudra, A. Kumar, and N. R. Moparthi, “AESPNet: Attention Enhanced Stacked Parallel Network to improve automatic Diabetic Foot Ulcer identification,” Image Vis. Comput., vol. 138, 2023, doi: 10.1016/j.imavis.2023.104809.

L. Alzubaidi, A. A. Abbood, M. A. Fadhel, O. Al-Shamma, and J. Zhang, “Comparison of hybrid convolutional neural networks models for diabetic foot ulcer classification,” Journal of Engineering Science and Technology, vol. 16, no. 3, 2021.

F. Arnia, K. Saddami, R. Roslidar, R. Muharar, and K. Munadi, “Towards accurate Diabetic Foot Ulcer image classification: Leveraging CNN pre-trained features and extreme learning machine,” Smart Health, vol. 33, Sep. 2024, doi: 10.1016/j.smhl.2024.100502.

E. Santos, F. Santos, J. Dallyson, K. Aires, J. M. R. S. Tavares, and R. Veras, “Diabetic Foot Ulcers Classification using a fine-tuned CNNs Ensemble,” in Proceedings - IEEE Symposium on Computer-Based Medical Systems, 2022. doi: 10.1109/CBMS55023.2022.00056.