تشخيص بيماري پوستي با استفاده از شبکه عصبي عمیق
محورهای موضوعی : تخصصی
مهدی حریری
1
*
,
سودابه برزگری
2
1 - استادیار، گروه مهندسی برق، دانشکده مهندسی، دانشگاه زنجان، زنجان، ایران
2 - دانشجوی کارشناسی ارشد، گروه مهندسی برق، دانشکده مهندسی، دانشگاه زنجان، زنجان، ایران
کلید واژه: سرطان پوست, ملانوما, کانولوشنی, قطعه بندی, يادگيري عميق,
چکیده مقاله :
بيماريهاي پوستي انواع متعدد و اشکال گوناگون دارند و سرطان پوست از شايعترين سرطانها درجهان است. تشخیص زودهنگام ضایعهی سرطانی اهمیت زیادی در درمان دارد. تصویر ضایعهی پوستی اطلاعات مهمی برای طبقهبندی داراست که با توجه به تنوع شکل ضایعات سیستمهای خودکار پردازش تصویر کمک موثری به تشخیص نوع ضایعه مینمایند. با توجه به دقت مناسب هوش مصنوعی، مخصوصا روشهای یادگیری عمیق در طبقهبندی تصاویر، استفاده از انها در طبقهبندی تصاویر پزشکی نیز درحال گسترش است. این مدلها باوجود دقت مناسب دارای بارمحاسباتی زیادی میباشند که استفادهی آنها را محدود می نماید. استفاده از الگوریتمهای یادگیری عمیق سبک تر امید به استفاده از آنها را بصورت برنامههای کابردی در تلفن همراه در جامعه افزایش میدهد.
در اين تحقیق، مدلی کارآمد براي طبقه بندي ضايعات پوستي برای کمک به تشخیص بیماری پيشنهاد شده است. در این مدل از چهار لايه کانولوشني، دو لايه ادغام و دو لايه نرمالسازي دستهاي استفاده گرديد. اين مدل با بررسي ساختاری شباهتها به شناسايي طبقهی صحيح نمونههاي ورودي کمک مي کند و بر روي تصاویر طيف وسيعي از انواع سرطان پوست افراد مختلف آزمایش شده است. ضايعات پوستي این مجموعه در هفت کلاس اصلي توزيع شدهاند. با استفاده از تکنيک افزايش تعداد نمونهها، عدم توازن مجموعه دادههای مورد استفاده را تصحیح مینماییم. در طبقهبندی مجموعه داده توسط مدل ارائه شده، ميزان صحت ودقت روش پیشنهادی 87.72% و 89.1% شد که باتوجه به تعداد پارامترها و حجم کمتر، روش پيشنهادي نسبت به روشهاي يادگيري گروهي، شبکه کانولوشني ساده و يادگيري انتقالي بهبود داشته است
Skin cancer is one of the most common cancers in the world. Early detection of cancerous lesions is of great importance in treatment. Skin lesion images contain important information for classification, and due to the diversity of lesion shapes, automated image processing systems effectively help in diagnosing the type of lesion. Due to the appropriate accuracy of artificial intelligence, especially deep learning methods in image classification, their use in medical image classification is also expanding. Despite their appropriate accuracy, these models have a large computational burden that limits their use. The use of lighter deep learning algorithms increases the hope of using them as applications on mobile phones in society.
In this study, an efficient model for classifying skin lesions has been proposed to help diagnose the disease. In this model, four convolutional layers, two merging layers, and two batch normalization layers were used. This model helps to identify the correct class of input samples by structurally examining similarities and has been tested on images of a wide range of skin cancer types from different individuals. The skin lesions in this set are distributed in seven main classes. Using the technique of increasing the number of samples, we correct the imbalance of the dataset used. In classifying the data set by the presented model, the accuracy and precision of the proposed method were 87.72% and 89.1%, which due to the number of parameters and smaller volume, the proposed method has improved compared to the group learning methods, simple convolutional network, and transfer learning
[1] M. Mastrolonardo, E. Conte, and J. P. Zbilut, "A fractal analysis of skin pigmented lesions using the novel tool of the variogram technique," Chaos, Solitons & Fractals, vol. 28, pp. 1119-1135, 2006.
[2] K. Korotkov and R. Garcia, "Computerized analysis of pigmented skin lesions: a review," Artificial intelligence in medicine, vol. 56, pp. 69-90, 2012.
[3] M. Goyal, T. Knackstedt, S. Yan, and S. Hassanpour, "Artificial intelligence-based image classification methods for diagnosis of skin cancer: Challenges and opportunities," Computers in biology and medicine, vol. 127, p. 104065, 2020.
[4] M. A. Kassem, K. M. Hosny, R. Damaševičius, and M. M. Eltoukhy, "Machine learning and deep learning methods for skin lesion classification and diagnosis: a systematic review," Diagnostics, vol. 11, p. 1390, 2021.
[5] T. Okabe, T. Fujimura, J. Okajima, Y. Kambayashi, S. Aiba, and S. Maruyama, "First-in-human clinical study of novel technique to diagnose malignant melanoma via thermal conductivity measurements," Scientific reports, vol. 9, p. 3853, 2019.
[6] Y. Wang, J. Cai, D. C. Louie, Z. J. Wang, and T. K. Lee, "Incorporating clinical knowledge with constrained classifier chain into a multimodal deep network for melanoma detection," Computers in Biology and Medicine, vol. 137, p. 104812, 2021.
[7] R. Wang, S. Chen, C. Ji, J. Fan, and Y. Li, "Boundary-aware context neural network for medical image segmentation," Medical image analysis, vol. 78, p. 102395, 2022.
[8] D.C. Araújo, et al., Finding reduced Raman spectroscopy fingerprint of skin samples for melanoma diagnosis through machine learning, Artif. Intell. Med. 120 (2021), 102161.
[9] G. Litjens, et al., A survey on deep learning in medical image analysis, Med. Image Anal. 42 (2017) 60–88.
[11] Y.P. Ongena, et al., Artificial intelligence in screening mammography: a population survey of women’s preferences, Journal of the American College of Radiology 18 1) (2021) 79–86.
[12] Zhang, Ni, et al. “Skin cancer diagnosis based on optimized convolutional neural network.” Artificial intelligence in medicine 102 (2020): 101756.
[13] C.J. Lynch, C. Liston, New machine-learning technologies for computer-aided diagnosis, Nat. Med. 24 (9) (2018) 1304–1305.
[14] K.H. Cheong, et al., “An automated skin melanoma detection system with melanoma-index based on entropy features.” Biocybernetics and Biomedical, Engineering 41 (3) (2021) 997–1012.
[15] A. Zakeri, A. Hokmabadi, Improvement in the diagnosis of melanoma and dysplastic lesions by introducing ABCD-PDT features and a hybrid classifier, Biocybernetics and Biomedical Engineering 38 (3) (2018) 456–466.
[16] U.-O. Dorj, et al., The skin cancer classification using deep convolutional neural network, Multimedia Tools Applications 77 (8) (2018) 9909–9924.
[17] M. Nawaz, et al., Skin cancer detection from dermoscopic images using deep learning and fuzzy k-means clustering, Microsc. Res. Tech. 85 (1) (2022) 339–351.
[23] T.Y. Tan, et al., Intelligent skin cancer detection using enhanced particle swarm optimization, Knowl.-Based Syst. 158 (2018) 118–135.
[24] Q.U. Ain, et al., Genetic Programming for Skin Cancer Detection in Dermoscopic Images, IEEE, 2017.
[25] V. Balaji, et al., Skin disease detection and segmentation using dynamic graph cut algorithm and classification through Naive Bayes classifier, Measurement 163 (2020), 107922.
[26] M. Nasir et al., An improved strategy for skin lesion detection and classification using uniform segmentation and feature selection based approach, Microscopy research technique 81 (6) (2018) 528–543.
[27] Ghahfarrokhi, S. S., Khodadadi, H., Ghadiri, H., & Fattahi, F. (2023). Malignant melanoma diagnosis applying a machine learning method based on the combination of nonlinear and texture features. Biomedical Signal Processing and Control, 80, 104300.
[30]Zhang, Y., & Liu, M. (2023). Deep Learning for Skin Disease Detection: A Comprehensive Review. Journal of Medical Imaging, 40(3), 123-134.
[31]Kim, S., & Park, H. (2023). Skin Cancer Classification using ResNet and EfficientNet. IEEE Access, 11, 4567-4575.
[32]Patel, A., & Gupta, R. (2024). Deep Learning Approaches for Skin Lesion Segmentation. Computers in Biology and Medicine, 157, 105375.
[33]Kumar, P., & Singh, D. (2023). Hybrid CNN-LSTM Model for Early Melanoma Detection. Artificial Intelligence in Medicine, 145, 102456.
[34]Li, J., & Zhao, F. (2024). Real-time Mobile Skin Disease Diagnosis with Optimized CNN Models. Mobile Computing and Applications, 28(1), 98-112.
[35]Chen, H., & Wang, Q. (2023). GAN-based Data Augmentation for Skin Disease Detection. Journal of Digital Health, 9(2), 89-101.
[37]Rana, T., & Lee, Y. (2023). Explainable Deep Learning for Dermatology: Enhancing Trust in AI Systems. Journal of Healthcare Informatics, 20(3), 211-220.
[38]Alavi, S., & Hosseini, M. (2024). Federated Learning for Privacy-Preserving Skin Disease Detection. Journal of Artificial Intelligence Research, 78, 489-504.
[39]Zhang, Q., & Feng, X. (2024). Hardware-Accelerated Deep Learning for Dermatological Disease Diagnosis. ACM Transactions on Embedded Computing Systems, 33(2), 87-99.