MULTITASK LEARNING UNTUK ANALISIS LUKA BAKAR MENGGUNAKAN U-NET, EFFICIENTNETB4, DAN VISION TRANSFORMER
Abstract
Luka bakar merupakan cedera serius yang dapat mengancam jiwa dan memerlukan penanganan cepat sesuai tingkat keparahannya. Keterbatasan tenaga medis terlatih di daerah terpencil mendorong pengembangan sistem diagnosis otomatis berbasis kecerdasan buatan. Penelitian ini mengembangkan sistem analisis luka bakar berbasis deep learning dengan pendekatan MultiTask Learning (MTL) yang mampu melakukan segmentasi area luka, klasifikasi derajat keparahan (I, II, dan III), serta estimasi rasio area luka secara bersamaan dalam satu model. Tiga arsitektur dibandingkan, yaitu U-Net, EfficientNetB4, dan Vision Transformer (ViT), menggunakan 1.846 citra luka bakar dari Roboflow dan Kaggle. Dataset dibagi menjadi data latih, validasi, dan uji dengan rasio 80:10:10 menggunakan stratified sampling. Hasil menunjukkan bahwa U-Net memberikan performa segmentasi terbaik dengan nilai Dice 0,6773 dan IoU 0,5121. EfficientNetB4 mencapai akurasi klasifikasi tertinggi sebesar 77,30%, sedangkan Vision Transformer (ViT) unggul pada estimasi rasio area luka dengan MAE 0,0819 dan R² 0,5107. Temuan ini menunjukkan bahwa tidak ada satu arsitektur yang unggul pada semua tugas, sehingga pemilihan model perlu disesuaikan dengan prioritas klinis. Sistem ini berpotensi mendukung diagnosis luka bakar di fasilitas kesehatan dengan sumber daya terbatas.
Downloads
References
R. Niri dkk., "Wound Segmentation with U-Net Using a Dual Attention Mechanism and Transfer Learning," Journal of Imaging Informatics in Medicine, vol. 38, no. 5, pp. 3351–3365, Oct. 2025.
World Health Organization, "Burns," WHO Fact Sheet, Oct. 2023. [Online]. Available: https://www.who.int/news-room/fact-sheets/detail/burns.
B. Cassidy, C. Kendrick, N. D. Reeves, J. M. Pappachan, and M. H. Yap, "Deep learning in chronic wound segmentation: a comprehensive review and meta-analysis," The Visual Computer, vol. 41, pp. 11885–11908, 2025, doi: 10.1007/s00371-025-04133-y.
H. He, C. Sun, Z. Wang, and Y. Dan, "Linear Transformer Based U-Shaped Lightweight Segmentation Network," Journal of Advanced Computational Intelligence and Intelligent Informatics, vol. 29, no. 6, pp. 1319–1328, 2025.
X. Li, Z. Liu, and L. Liu, "Pediatric BurnNet: Robust Multi-Class Segmentation and Severity Recognition under Real-World Imaging Conditions," SAGE Open Medicine, vol. 13, pp. 1–12, 2025, doi: 10.1177/20503121251360090.
R. Zhang, D. Tian, D. Xu, W. Qian, and Y. Yao, "A Survey of Wound Image Analysis Using Deep Learning: Classification, Detection, and Segmentation," IEEE Access, vol. 10, 2022, doi: 10.1109/ACCESS.2022.3194529.
Y. Gao, Y. Jiang, Y. Peng, F. Yuan, X. Zhang, and J. Wang, "Medical Image Segmentation: A Comprehensive Review of Deep Learning-Based Methods," Tomography, vol. 11, no. 5, May 2025, doi: 10.3390/tomography11050052.
S. Minaee, Y. Boykov, F. Porikli, A. Plaza, N. Kehtarnavaz, and D. Terzopoulos, "Image Segmentation Using Deep Learning: A Survey," IEEE Trans. Pattern Anal. Mach. Intell., vol. 44, no. 7, pp. 3523–3542, Jul. 2022.
K. Chtouki dkk., "Multi-Task Deep Learning for Simultaneous Classification and Segmentation of Cancer Pathologies in Diverse Medical Imaging Modalities," Onco, vol. 5, no. 3, Jul. 2025, doi: 10.3390/onco5030034.
Y. Zhang, J. Wang, J. M. Gorriz, and S. Wang, "Deep Learning and Vision Transformer for Medical Image Analysis," Journal of Imaging, vol. 9, no. 7, Jul. 2023, doi: 10.3390/jimaging9070147.
A. Zhang, Z. C. Lipton, M. Li, and A. J. Smola, Dive into Deep Learning. Cambridge, UK: Cambridge University Press, 2023.
A. Abdelrahman and S. Viriri, "EfficientNet Family U-Net Models for Deep Learning Semantic Segmentation of Kidney Tumors on CT Images," Frontiers in Computer Science, vol. 5, 2023, doi: 10.3389/fcomp.2023.1235622.
R. A. B. Kusuma, B. Irawan, dan A. Khamid, "Klasifikasi Penyakit Kulit Wajah Menggunakan Convolutional Neural Network EfficientNet-B3," JITET (Jurnal Informatika dan Teknik Elektro Terapan), vol. 14, no. 1, 2026, doi: 10.23960/jitet.v14i1.8721.
L. Maurya and S. Mirza, "MCTFWC: a multiscale CNN-transformer fusion-based model for wound image classification," Signal, Image and Video Processing, May 2025, doi: 10.1007/s11760-025-04072-0.
R. R. Ar, Agusriyati, dan S. Moka, "Pendeteksian Dini Stunting pada Balita Menggunakan Vision Transformer (ViT) Berbasis Citra Tubuh," JITET (Jurnal Informatika dan Teknik Elektro Terapan), vol. 13, no. 3S1, 2025, doi: 10.23960/jitet.v13i3S1.7888.
A. Kendall, Y. Gal, and R. Cipolla, "Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics," in Proc. IEEE/CVF Conf. Computer Vision and Pattern Recognition (CVPR), Salt Lake City, USA, Jun. 2018, pp. 7482–7491.
R. Szeliski, Computer Vision: Algorithms and Applications, 2nd ed. Cham, Switzerland: Springer Nature, 2022.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
