CNN tabanlı derin öğrenme yaklaşımı ile kanser görüntülerinin sınıflandırılması

Year 2023, , 30 - 38, 15.01.2023

Halit Çetiner

https://doi.org/10.28948/ngumuh.1143693

Abstract

Halk arasında melanoma (mel), dermatofibroma (df), ve vascular (vasc), bening keratosis (bkl), melanocytic nevi (nv), basal cell carcinoma (bcc), actinic keratosis (akiec) olarak bilinen cilt kanserleri yüksek benzerliğe sahiptir. Belirtilen cilt kanserlerinin erken aşamada doğru bir şekilde sınıflandırılması insan yaşamını kurtarması açısından önemlidir. Bu makalede yaygın görülen cilt kanserlerinin sınıflandırma süreçleri için yüksek doğruluklu bir derin öğrenme modeli önerilmiştir. Önerilen model, genel olarak iş yoğunluğu yüksek olan cilt uzmanlarına yardımcı, hızlı tanı ve sınıflandırma yetkinliğine sahip bir modeldir. Birbirine oldukça benzer olan cilt kanserlerinin sınıflandırılmasında, swish ve ReLU aktivasyon fonksiyonlarının avantajlarından faydalanan 30 katmanlı bir CNN modeli önerilmiştir. Bu model kullanılarak akiec, bcc, bkl, df, nv, vasc, mel adlı cilt kanserlerinin sınıflandırılmasında sırasıyla 0.99%, 0.99%, 0.96%, 0.99%, 0.92%, 0.99%, 0.95% F1 score değerleri elde edilmiştir. Akiec, bcc, bkl, df, nv, vasc, mel adlı cilt kanserlerinin sınıflandırılmasında precision ve recall ölçüm metrikleri açısından sırasıyla 0.99%, 0.99, 0.93, 0.99, 0.97, 0.99, 0.94 precision ve 0.99, 0.98, 0.99, 1, 0.87, 1, 0.97 recall değerleri elde edilmiştir. Elde edilen performans sonuçlarına göre önerilen modelin birbirine oldukça benzer yedi farklı cilt kanserini doğru bir şekilde sınıflandırdığı söylenebilir.

Keywords

CNN, Rastgele yeniden örneklendirme algoritması, Cilt kanseri, Derin öğrenme

Classification of cancer images with CNN-based deep learning approach

Abstract

Skin cancers known as melanoma (mel), dermatofibroma (df), and vascular (vasc), benign keratosis (bkl), melanocytic nevi (nv), basal cell carcinoma (bcc), actinic keratosis (akiec) have a high similarity. Accurate classification of specified skin cancers at an early stage is important in terms of saving human life. In this article, a high-accuracy deep learning model is proposed for the classification processes of common skin cancers. The proposed model is a model that helps skin specialists with a high workload and has rapid diagnosis and classification competence. A 30-layer CNN model is proposed that takes advantage of the swish and ReLU activation functions in the classification of highly similar skin cancers. Using this model, 0.99%, 0.99%, 0.96%, 0.99%, 0.92%, 0.99%, 0.95% F1 score values were obtained in the classification of skin cancers named akiec, bcc, bkl, df, nv, vasc, mel, respectively. In terms of precision and recall measurement metrics in the classification of skin cancers named Akiec, bcc, bkl, df, nv, vasc, mel, respectively, 0.99%, 0.99, 0.93, 0.99, 0.97, 0.99, 0.94 precision and 0.99, 0.98, 0.99, 1, 0.87, 1, 0.97 recall values were obtained. Based on the performance results obtained, it can be said that the proposed model correctly classifies seven very similar skin cancers.

Keywords

CNN, Random Over Sampler algorithm, Skin cancer, Deep learning

References

• H. Younis, M. H. Bhatti, and M. Azeem, Classification of Skin Cancer Dermoscopy Images using Transfer Learning, in 2019 15th International Conference on Emerging Technologies, 1–4. 2019. https://doi.org/10. 1109/ICET48972.2019.8994508.
• C. De Martel, J. Ferlay, S. Franceschi, J. Vignat, F. Bray, D. Forman, and M. Plummer, Global burden of cancers attributable to infections in 2008: a review and synthetic analysis, Lancet Oncol., 13(6), 607–615, 2012.
• R. Perroy, World population prospects, United Nations, 1(6042), 587–592, 2015.
• D. Pimentel, S. Cooperstein, H. Randell, D. Filiberto, S. Sorrentino, B. Kaye, C. Nicklin, J. Yagi, J. Brian, J. O'Hern, A. Habas, and Weinstein, Ecology of Increasing Diseases: Population Growth and Environmental Degradation, Hum. Ecol. Interdiscip. J., 35(6), 653–668, 2007. https://doi.org/10.1007/s10745-007-9128-3.
• N. Bruce, R. Perez-Padilla, and R. Albalak, The health effects of indoor air pollution exposure in developing countries, Geneva World Heal. Organ., 11, 2002.
• U.-O. Dorj, K.-K. Lee, J.-Y. Choi, and M. Lee, The skin cancer classification using deep convolutional neural network, Multimed. Tools Appl., 77(8), 9909–9924, 2018. https://doi.org/10.1007/s11042-018-5714-1.
• K. E. Kim, D. Cho, and H. J. Park, Air pollution and skin diseases: Adverse effects of airborne particulate matter on various skin diseases, Life Sci., 152, 126–134, 2016.
• A. J. McMichael and T. McMichael, Planetary overload: global environmental change and the health of the human species. Cambridge University Press, 1993.
• P. Martens and A. J. McMichael, Environmental change, climate and health: issues and research methods. Cambridge University Press, 2009.
• R. L. McKenzie, L. O. Björn, A. Bais, and M. Ilyasd, Changes in biologically active ultraviolet radiation reaching the Earth’s surface, Photochem. Photobiol. Sci., 2(1), 5–15, 2003.
• F. W. Alsaade, T. H. H. Aldhyani, and M. H. Al-Adhaileh, Developing a Recognition System for Diagnosing Melanoma Skin Lesions Using Artificial Intelligence Algorithms., Comput. Math. Methods Med., 9998379, 2021. https://doi.org/10.1155/2021/99 98379.
• D. B. Mendes and N. C. da Silva, Skin lesions classification using convolutional neural networks in clinical images, arXiv Prepr. arXiv1812.02316, 2018.
• Y. LeCun, Y. Bengio, and G. Hinton, Deep learning, Nature, 521(7553), 436–444, 2015. https://doi.org/10 .1038/nature14539.
• M. Choudhary, S. S. Chouhan, E. S. Pilli, and S. K. Vipparthi, BerConvoNet: A deep learning framework for fake news classification, Appl. Soft Comput., 110, 107614, 2021. https://doi.org/10.1016/j.asoc.2021.107 614.
• T. Chen, R. Xu, Y. He, and X. Wang, Improving sentiment analysis via sentence type classification using BiLSTM-CRF and CNN, Expert Syst. Appl., 72, 221–230, 2017. https://doi.org/10.1016/j.eswa.2016.10 .065.
• X. Xu, L. Zhang, J. Li, Y. Guan, and L. Zhang, A Hybrid Global-Local Representation CNN Model for Automatic Cataract Grading, IEEE J. Biomed. Heal. Informatics, 24(2), 556–567, 2020, https://doi.org/ 10.1109/JBHI.2019.2914690.
• A. Nabil, M. Seyam, and A. Abou-Elfetouh, Deep Neural Networks for Predicting Students’ Performance, in Proceedings of the 52nd ACM Technical Symposium on Computer Science Education, 2021. https://doi.org/10.1145/3408877.3439685.
• J. Liu, K. Li, B. Song, and L. Zhao, A Multi-stream Convolutional Neural Network for Micro-expression Recognition Using Optical Flow and {EVM}, CoRR, 2020.
• W. Lu, H. Hou, and J. Chu, Feature fusion for imbalanced ECG data analysis, Biomed. Signal Process. Control, 41, 152–160, 2018. https://doi.org/10 .1016/j.bspc.2017.11.010.
• B. Titus Josef, H. Achim, U. Jochen Sven, G. Niels, S. Dirk, K. Joachim, B. Carola, S. Theresa, E. Alexander , and V. Christof, Skin cancer classification using convolutional neural networks: systematic review, J. Med. Internet Res., 20(10), 2018.
• K. M. Hosny, M. A. Kassem, and M. M. Fouad, “Classification of Skin Lesions into Seven Classes Using Transfer Learning with AlexNet, J. Digit. Imaging, 33(5), 1325–1334, 2020. https://doi.org/ 10.1007/s10278-020-00371-9.
• L. Yu, H. Chen, Q. Dou, J. Qin, and P.-A. Heng, Automated Melanoma Recognition in Dermoscopy Images via Very Deep Residual Networks, IEEE Trans. Med. Imaging, 36(4), 994–1004, 2017. https://doi.org/ 10.1109/TMI.2016.2642839.
• M. Ramachandro, T. Daniya, and B. Saritha, Skin Cancer Detection Using Machine Learning Algorithms, in 2021 Innovations in Power and Advanced Computing Technologies, , 1–7, 2021. https://doi.org/ 10.1109/i-PACT52855.2021.9696874.
• W. Sae-Lim, W. Wettayaprasit, and P. Aiyarak, Convolutional Neural Networks Using MobileNet for Skin Lesion Classification, in 2019 16th International Joint Conference on Computer Science and Software Engineering,. 242–247, 2019. https://doi.org/10.1109 /JCSSE.2019.8864155.
• A. M. Alhassan and W. M. N. W. Zainon, Brain tumor classification in magnetic resonance image using hard swish-based RELU activation function-convolutional neural network, Neural Comput. Appl., 33(15), 9075–9087, 2021. https://doi.org/10.1007/s00521-020-0567 1-3.
• P. Tschandl, C. Rosendahl, and H. Kittler, The HAM10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions, Sci. data, 5(1), 1–9, 2018.
• S. Qian, C. Ning, and Y. Hu, MobileNetV3 for Image Classification, in 2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering, 490–497, 2021. https:// doi.org/10.1109/ICBAIE52039.2021.9389905.
• K. Eckle and J. Schmidt-Hieber, A comparison of deep networks with ReLU activation function and linear spline-type methods, Neural Networks, 110, 232–242, 2019. https://doi.org/10.1016/j.neunet.2018.11.005.
• G. Lin and W. Shen, Research on convolutional neural network based on improved Relu piecewise activation function, Procedia Comput. Sci., 131, 977–984, 2018. https://doi.org/10.1016/j.procs.2018.04.239.
• Y. Yu, K. Adu, N. Tashi, P. Anokye, X. Wang, and M. A. Ayidzoe, RMAF: Relu-Memristor-Like Activation Function for Deep Learning, IEEE Access, 8, 72727–72741, 2020. https://doi.org/10.1109/ACCESS.2020.2 987829.
• S. Rubinstein-Salzedo, Big o notation and algorithm efficiency, in Cryptography, Springer, 75–83, 2018.
• S.-R.-S. Jianu, L. Ichim, D. Popescu, and O. Chenaru, Advanced Processing Techniques for Detection and Classification of Skin Lesions, in 2018 22nd International Conference on System Theory, Control and Computing, 498–503, 2018. https://doi.org/10.11 09/ICSTCC.2018.8540732.
• I. Giotis, N. Molders, S. Land, M. Biehl, M. F. Jonkman, and N. Petkov, MED-NODE: A computer-assisted melanoma diagnosis system using non-dermoscopic images, Expert Syst. Appl., 42(19), 6578–6585, 2015.
• J. Kawahara, A. BenTaieb, and G. Hamarneh, Deep features to classify skin lesions, in 2016 IEEE 13th international symposium on biomedical imaging, 1397–1400, 2016. https://doi.org/10.1109/ISBI.2016.7 493528.