ARSITEKTUR ENSEMBLE U-NET UNTUK SEGMENTASI KANKER PAYUDARA OTOMATIS PADA GAMBAR MAMMOGRAM
Abstract
Kanker payudara masih menjadi salah satu penyebab utama kematian akibat kanker pada wanita di seluruh dunia. Deteksi dini melalui skrining rutin menggunakan mammogram terbukti efektif dalam mengurangi angka kematian. Namun, interpretasi mammogram secara manual memerlukan waktu, bersifat subjektif, dan sering kali membutuhkan radiolog yang berpengalaman. Untuk mengatasi tantangan ini, penelitian ini mengusulkan arsitektur Ensemble U-Net untuk melakukan segmentasi kanker payudara secara otomatis pada citra mammogram. Proses segmentasi melibatkan beberapa langkah, termasuk praproses (penghapusan latar belakang, penghapusan otot pektoral, peningkatan kontras, dan pengubahan ukuran), dilanjutkan dengan segmentasi menggunakan ensemble model: Inception V3-U-Net, ResNet50-U-Net, VGG19-U-Net, dan U-Net kustom. Segmentasi akhir dicapai dengan menggunakan voting soft dan filter Gaussian 2D untuk mereduksi noise, diikuti dengan thresholding untuk segmentasi biner. Pendekatan ensemble menunjukkan peningkatan akurasi segmentasi dengan menggabungkan kekuatan dari beberapa model U-Net. Kinerja model dievaluasi menggunakan metrik seperti akurasi, sensitivitas, spesifisitas, koefisien Dice, dan Intersection over Union (IoU). Hasil eksperimen menunjukkan bahwa Ensemble U-Net memiliki kinerja yang lebih baik dibandingkan dengan model individu, terutama pada citra mammogram yang kompleks.
Full Text:
PDF (Bahasa Indonesia)References
M. A. Bashar and N. Begam, “Breast cancer sur-passes lung cancer as the most commonly diagnosed cancer worldwide,” Indian Journal of Cancer, vol. 59, no. 3, pp. 438–439, 2022.
T. Hovda, K. Tsuruda, S. R. Hoff, K. K. Sahlberg, and S. Hofvind, “Radiological review of prior screening mammograms of screen-detected breast cancer,” Eu-ropean radiology, vol. 31, pp. 2568–2579, 2021.
D. Schopper and C. de Wolf, “How effective are breast cancer screening programmes by mammogra-phy? Review of the current evidence,” European journal of cancer, vol. 45, no. 11, pp. 1916–1923, 2009.
O. Ronneberger, P. Fischer, and T. Brox, “U-net: Con-volutional networks for biomedical image segmenta-tion,” Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9351, pp. 234–241, 2015, doi: 10.1007/978-3-319-24574-4_28.
X. Yuan, L. Xie, and M. Abouelenien, “A regularized ensemble framework of deep learning for cancer de-tection from multi-class, imbalanced training data,” Pattern Recognition, 2018, doi: 10.1016/j.patcog.2017.12.017.
S. Wang, Y. Yin, G. Cao, B. Wei, Y. Zheng, and G. Yang, “Hierarchical retinal blood vessel segmentation based on feature and ensemble learning,” Neurocom-puting, vol. 149, no. PB, pp. 708–717, 2015, doi: 10.1016/j.neucom.2014.07.059.
H. Li et al., “Fully convolutional network ensembles for white matter hyperintensities segmentation in MR images,” NeuroImage, vol. 183, no. July, pp. 650–665, 2018, doi: 10.1016/j.neuroimage.2018.07.005.
X. Xia, C. Xu, and B. Nan, “Inception-v3 for flower classification,” in 2017 2nd international conference on image, vision and computing (ICIVC), IEEE, 2017, pp. 783–787.
I. Z. Mukti and D. Biswas, “Transfer learning based plant diseases detection using ResNet50,” in 2019 4th International conference on electrical information and communication technology (EICT), IEEE, 2019, pp. 1–6.
A. Sengupta, Y. Ye, R. Wang, C. Liu, and K. Roy, “Going deeper in spiking neural networks: VGG and residual architectures,” Frontiers in neuroscience, vol. 13, p. 95, 2019.
DOI: http://dx.doi.org/10.36564/njca.v9i2.378
Refbacks
- There are currently no refbacks.
Copyright (c) 2024 Cahya Bagus Sanjaya

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.