U-Net-RCB7: Image Segmentation Algorithm

Yıl 2023, Cilt: 26 Sayı: 4, 1555 - 1562, 01.12.2023

Cihan Akyel , Nursal Arıcı

https://doi.org/10.2339/politeknik.1208936

Cited By: 2

Öz

The incidence of skin cancer is increasing. Early detection of cases of skin cancer is vital for treatment. Recently, computerized methods have been widely used in cancer diagnosis. These methods have important advantages such as no human error, short diagnosis time, and low cost. We can segment skin cancer images using deep learning and image processing. Properly segmented images can help doctors predict the type of skin cancer. However, skin images can contain noise such as hair. These noises affect the accuracy of segmentation. In our study, we created a noise dataset. It contains 3000 images and masks. We performed noise removal and lesion segmentation by utilizing the ISIC and PH2. We have developed a new deep learning model called U-Net-RCB7. U-Net-RCB7 contains EfficientNetB7 as the encoder and ResNetC before the last layer. This paper uses a modified U-Net model. Images were divided into 36 layers to prevent loss of pixel values in the images. As a result, noise removal and lesion segmentation were 96% and 98.36% successful, respectively.

Anahtar Kelimeler

deep learning, skin cancer, devision support systems, image proccessing, UNet

U-Net-RCB7: Görüntü Bölütleme Algoritması

Öz

Cilt kanseri insidansı artmaktadır. Cilt kanseri vakalarının erken tespiti tedavi için hayati önem taşır. Son zamanlarda kanser teşhisinde bilgisayarlı yöntemler yaygın olarak kullanılmaktadır. Bu yöntemlerin insan hatası olmaması, kısa teşhis süresi ve düşük maliyet gibi önemli avantajları vardır. Derin öğrenme ve görüntü işlemeyi kullanarak cilt kanseri görüntülerini segmentlere ayırabiliriz. Düzgün şekilde bölümlere ayrılmış görüntüler, doktorların cilt kanseri türünü tahmin etmesine yardımcı olabilir. Bununla birlikte, cilt görüntüleri saç gibi gürültüler içerebilir. Bu sesler, segmentasyonun doğruluğunu etkiler. Çalışmamızda bir gürültü veri seti oluşturduk. 3000 resim ve maske içerir. ISIC ve PH2'yi kullanarak gürültü giderme ve lezyon segmentasyonu gerçekleştirdik. U-Net-RCB7 adlı yeni bir derin öğrenme modeli geliştirdik. U-Net-RCB7, kodlayıcı olarak EfficientNetB7'yi ve son katmandan önce ResNetC'yi içerir. Bu yazıda değiştirilmiş bir U-Net modeli kullanılmaktadır. Görüntülerde piksel değerlerinin kaybolmaması için görüntüler 36 katmana ayrılmıştır. Sonuç olarak, gürültü giderme ve lezyon segmentasyonu sırasıyla %96 ve %98.36 başarılı olmuştur.

Anahtar Kelimeler

derin öğrenme, cilt kanseri, karar destekleme sistemleri, UNet

Kaynakça

• [1] Thapar, P., Rakhra, M., Cazzato, G., Hossain, S.; ”A Novel Hybrid Deep Learning Approach for Skin Lesion Segmentation and Classification”, Hindawi Journal of Healthcare Engineering, 2022: 1-21, (2022).
• [2] Siegel, R. L., Miller, K. D., and Jemal, A., “Cancer statistics”, CA: A Cancer Journal of Clinicians, 1: 7-33, (2021).
• [3] Unver, H. M., and Ayan, E. “Skin Lesion Segmentation in Dermoscopic Images with Combination of YOLO and GrabCut Algorithm”, Diagnostics Journal, 9(3): 1-21, (2019).
• [4] A. Kumar, A. Vatsa, ”Untangling Classification Methods for Melanoma Skin Cancer,” Front. Big Data, 5, (2022).
• [5] A. A, Alfi, M. Rahman, M. Shorfuzzaman, A. Nazir, ”A Non-Invasive Interpretable Diagnosis of Melanoma Skin Cancer Using Deep Learning and Ensemble Stacking of Machine Learning Models,” MDPI Diagnostic, 12(13):1-18, (2022).
• [6] W. Salma, A. S. Eltrass, ”Automated deep learning approach for classification of malignant melanoma and benign skin lesions,” Multimedia Tools and Applications, 2022. in Proc. The 36th International Conference on Machine Learning, California, USA, 9-15, (2019).
• [7] M. Kahia, A. Echtioui, F. Kallel, A. B. Hamida, ”Skin Cancer Classification using Deep Learning Models,” in Proc. International Conference on Agents and Artificial Intelligence, 554-559, (2022).
• [8] M. Arif, F. Philip, F. Ajesh, D. Izdrui, M. D. Craciun, O. Geman, ”Automated Detection of Nonmelanoma Skin Cancer Based on Deep Convolutional Neural Network,” Hindawi Journal of Healthcare Engineering, (2022).
• [9] I. Abunadi, E. M. Senan, ”Deep Learning and Machine Learning Techniques of Diagnosis Dermoscopy Images for Early Detection of Skin Diseases,” MDPI Electronics, 10(24):1-50, (2021).
• [10] E. U¨ nlu¨, E. C¸ ınar, ”Segmentation of Benign and Malign lesions on skin images using U-Net,” 2021 International Conference on Innovation and Intelligence for Informatics, Computing, and Technologies (3ICT), Zallaq, Bahrain, 165-169, 29-30 Sept. (2021).
• [11] L. Wei, N. J. R. Alex, T. Tardi and Z. Zhemin, “Digital hair removal by deep learning for skin lesion segmentation,” Pattern Recognition, 117: 1-15, (2021).
• [12] K. Zafar, S. O. Gilani, A. Waris, A. Ahmed, M. Jamil, A. S. Kashif and M. N. Khan, “Skin Lesion Segmentation from Dermoscopic Images Using Convolutional Neural Network,” Sensors Journal, 20(6): 1-14, (2020).
• [13] G. Zhang, X. Shen, S. Chen, L. Liang, Y. Luo, J. Yu And J. Lu, “DSM: A Deep Supervised Multi-Scale Network Learning for Skin Cancer Segmentation,” IEEE Access, 7:1-10, (2016).
• [14] K. Hasan, L. Dahal, P. N. Samarakoon, F. I. Tushara and R. Marti, “DSNet: Automatic Dermoscopic Skin Lesion Segmentation,” Computers in biology and medicine, 120: 426-434, (2020).
• [15] C. Akyel and N. Arıcı, “A New Approach to Hair Noise Cleaning and Lesion Segmentation in Images of Skin Cancer,” Journal of Polytechnic, 23(3): 821-828, (2020).
• [16] Y. Dong, L. Wang, S. Cheng and Y. Li, “FAC-Net: Feedback Attention Network Based on Context Encoder Network for Skin Lesion Segmentation,” Sensor Journal, 21(15): 1-17, (2021).
• [17] N. Sahin and N. Alpaslan, “Seg-Net Mimarisi Kullanılarak Cilt Lezyon B¨olu¨tleme Performansının İyileştirilmesi,” Avrupa Bilim ve Teknoloji Dergisi, special issue: 40-45, (2020).
• [18] P. Brahmbhatt and S. N. Rajan, “Skin Lesion Segmentation using Seg-Net with Binary CrossEntropy,” Vivechan International Journal of Research, 10(2): 22-31, (2019).
• [19] T. Phan, S. Kim, H. Yang and G. Lee, “Skin Lesion Segmentation by U-Net with Adaptive Skip Connection and Structural Awareness, ” Applied sciences, 11(10): 1-14, (2021).
• [20] F. Bagheri, M. J. Tarokh M. Ziaratban, “Skin lesion segmentation based on mask RCNN, Multi Atrous Full-CNN and A Geodesic Method,” International Journal of Imaging Systems and Technology, 31(3): 1609-1624, (2021).
• [21] C. Akyel, N. Arıcı, ”LinkNet-B7: Noise Removal and Lesion Segmentation in Images of Skin Cancer,” Mathematics, 10(5):736-751, (2022).
• [22] Peng Tang , Qiaokang Liang , Xintong Yan , Shao Xiang , Wei Sun , Dan Zhang , Gianmarc Coppola “Efficient skin lesion segmentation using separable-U-Net with stochastic weight averaging,” ELSEIVER Computer Methods and Programs in Biomedicine, 178: 289– 301, (2019).
• [23] C. Akyel, N. Arıcı, ”Hair Removal and Lesion Segmentation with FCN8- ResNetC and Image Processing in Images of Skin Cancer,” Journal of Information Technologies, 15(2), 231-238, (2022).
• [24] T. Mingxing, and V. L. Quoc, “EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks,” in Proc. The 36th International Conference on Machine Learning, California, USA, 9-15, (2019).
• [25] B. Baheti, S. Innani, S. Gajre and S. Talbar, “Eff-U-Net: A Novel Architecture for Semantic Segmentation in Unstructured Environment,” in Proc 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Seattle, USA, (2020).
• [26] J. Wang, X. Zhang, P. Lv, L. Zhou and H. Wang, “EAR-U-Net: EfficientNet and attention-based residual U-Net for automatic liver segmentation in CT,” arXiv, 1-26, (2021).
• [27] https://challenge2018.isic-archive.com/task3/training,/ “ISIC 2018”, (2023).
• [28] https://challenge.isic-archive.com/landing/2018/, “ISIC 2018”, (2023).
• [29] https://www.fc.up.pt/addi/PH2%20database.html3, “FCUP110”, (2023).
• [30] O. Ronneberger, P. Fischer and T. Brox, “U-Net: Convolutional Networks for Biomedical Image Segmentation,” in Proc. Medical Image Computing and Computer-Assisted Intervention– MICCAI 2015, Berlin, Germany, pp. 234-241, (2015).
• [31] M. Sandler, A. G. Howard, M. Zhu, A. Zhmoginov and L. Chen, “Mobilenetv2: Inverted residuals and linear bottlenecks,” in Proc.2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City,7USA, 4510–4520, (2018).
• [32] K. He, X. Zhang, S. Ren and J. Sun, “Deep Residual Learning for Image Recognition,” in Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, USA, (2016).
• [33] S. Shamim, M. J. Awan, A. M. Zain, U. Naseem, M. A. Mohammed and B. Garcia-Zapirain, ”Automatic COVID-19 Lung Infection Segmentation through Modified U-Net Model,” Journal of Healthcare Engineering, 2022(12):1-13, (2022).
• [34] D. Kingma, J. Ba, ”Adam: A Method for Stochastic Optimization,” Published as a conference paper at the 3rd International Conference for Learning Representations, San Diego, (2015).
• [35] R. Padilla, S. L. Netto, E. A. B. Da Silva, ”A Survey on Performance Metrics for Object-Detection Algorithms,” Conference: 2020 International Conference on Systems, Signals and Image Processing (IWSSIP), july, Puducherry, India,(2020).
• [36] P.Chatterjee, S. Jana and S. Ghosh, “Comparative Study of OpenCV Inpainting Algorithms,” Global Journal of Computer Science and Technology: G Interdisciplinary, 21 (2): 26-37, (2021).