Hibrit Aktarım Öğrenme Tekniklerini Kullanarak Beyin Tümörü Sınıflandırmasında Yeni Sonuçlar

Year 2025, Erken Görünüm, 1 - 1

Doğukan Kalender , Atahan Öztürk , Ömer Bilgin , Durmuş Özkan Şahin

https://doi.org/10.29109/gujsc.1501181

Abstract

Bu çalışmada derin öğrenme modelleri kullanılarak beyin MR görüntüleri işlenmiştir. Kullanılan veri seti tümör bulunan ve bulunmayan görüntülerden oluşmaktadır. Bu görüntüler bazı ön-işleme aşamalarından geçerek modellerin eğitimi için uygun hale getirilmektedir. Çalışmada kullanılan derin öğrenme modelleri aktarım öğrenmesine dayalı modellerden oluşmaktadır. Bunlar MobileNet, VGG19, DenseNet169, AlexNet, ResNet101 ve InceptionV3 modelleridir. Bu modeller önceden eğitilmiş yapıda bulunmaktadır. Bu nedenle derin öğrenme modellerinin daha iyi genelleme yaparak doğru tahminlerde bulunmasını sağlarlar. Modellerin performansını arttırmak için veri arttırma, erken durduma, öğrenme oranı zamanlayıcısı, katman dondurma ve özel katmanların eklenmesi gibi yöntemler kullanılmıştır. Yapılan deneylerde en yüksek başarım doğruluk metriğine göre %98.63 ile MobileNet modelinden elde edilmektedir. Daha sonra deneylerden elde edilen sonuçlara göre en başarılı üç modelin ikili kombinasyonları alınarak hibrit modeller oluşturulmuştur. Önerilen bu hibrit modellerin kullanılması ile elde edilen en yüksek başarım doğruluk metriğine göre %99.21’dir. Bu sonuç VGG19 ve DenseNet169 modellerinin birleştirilmesiyle elde edilmiştir. Tüm hibrit modellerden elde edilen sonuçlar göz önünde bulundurulduğunda sınıflandırma başarımında %2’den fazla iyileştirme sağlanmıştır.

Keywords

Derin öğrenme, beyin tümörü tespiti, aktarım öğrenmesi, görüntü işleme, tıbbi görüntü, tıbbi görüntü sınıflandırma

Ethical Statement

Çalışma, etik kurul izni veya herhangi bir özel izin gerektirmemektedir

Supporting Institution

Bu çalışma için herhangi bir kurum ve/veya kuruluştan destek alınmamıştır.

New Results in Brain Tumor Classification Using Hybrid Transfer Learning Techniques

Abstract

In this study, brain MRI images were processed using deep learning models. The data set used consists of images with and without tumor. These images are made suitable for training models by going through some pre-processing stages. The deep learning models used in the study consist of models based on transfer learning. These are MobileNet, VGG19, DenseNet169, AlexNet, ResNet101 and InceptionV3 models. These models are pre-trained. Therefore, they enable deep learning models to generalize better and make accurate predictions. Methods such as data augmentation, early stopping, learning rate timer, layer freezing and adding special layers have been used to increase the performance of the models. In the experiments, the highest performance is obtained from the MobileNet model with 98.63% according to the accuracy metric. Then, according to the results obtained from the experiments, hybrid models were created by taking binary combinations of the three most successful models. The highest performance achieved by using these proposed hybrid models is 99.21% according to the accuracy metric. This result was obtained by combining VGG19 and DenseNet169 models. Considering the results obtained from all hybrid models, more than 2% improvement in classification performance was achieved.

Keywords

Deep learning, brain tumor detection, transfer learning, image processing, medical image, medical image classification

References

• [1] Abd El Kader I, Xu G, Shuai Z, Saminu S, Javaid I, Salim Ahmad I. Differential deep convolutional neural network model for brain tumor classification. Brain Sciences. 2021; 11(3): 352.
• [2] Logeswari T, Karnan M. An improved implementation of brain tumor detection using segmentation based on hierarchical self organizing map. International Journal of Computer Theory and Engineering. 2010; 2(4): 591.
• [3] El-Dahshan ESA, Mohsen HM, Revett K, Salem AM. Computer-aided diagnosis of human brain tumor through MRI: A survey and a new algorithm. Expert Systems with Applications. 2014; 41(11): 5526–5545.
• [4] Chahal PK, Pandey S, Goel S. A survey on brain tumor detection techniques for MR images. Multimedia Tools and Applications. 2020; 79(29): 21771–21814.
• [5] Arı A, Alcin OF, Hanbay D. Brain MR image classification based on deep features by using extreme learning machines. Biomedical Journal of Scientific and Technical Research. 2020; 25(3).
• [6] Varshney S, Prajapati SK, Rajput S, Kaur M, Rakesh N, Goyal MK. Image processing based brain tumor detection. In: 2022 International Conference on Fourth Industrial Revolution Based Technology and Practices (ICFIRTP); 2022. p. 204–209. IEEE.
• [7] Anwar SM, Yousaf S, Majid M. Brain tumor segmentation on multimodal MRI scans using EMAP algorithm. In: 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC); 2018. p. 550–553. IEEE.
• [8] Methil AS. Brain tumor detection using deep learning and image processing. In: 2021 International Conference on Artificial Intelligence and Smart Systems (ICAIS); 2021. p. 100–108. IEEE.
• [9] Hashan AM, Agbozo E, Al-Saeedi AAK, Saha S, Haidari A, Rabi MNF. Brain tumor detection in MRI images using image processing techniques. In: 2021 4th International Symposium on Agents, Multi-Agent Systems and Robotics (ISAMSR); 2021. p. 24–28. IEEE.
• [10] Noori M, Bahri A, Mohammadi K. Attention-guided version of 2D UNet for automatic brain tumor segmentation. In: 2019 9th International Conference on Computer and Knowledge Engineering (ICCKE); 2019. p. 269–275. IEEE.
• [11] Sravanthi N, Swetha N, Devi PR, Rachana S, Gothane S, Sateesh N. Brain tumor detection using image processing. International Journal of Scientific Research in Computer Science, Engineering and Information Technology. 2021; 7(3): 348–352.
• [12] Malik M, Jaffar MA, Naqvi MR. Comparison of brain tumor detection in MRI images using straightforward image processing techniques and deep learning techniques. In: 2021 3rd International Congress on Human-Computer Interaction, Optimization and Robotic Applications (HORA); 2021. p. 1–6. IEEE.
• [13] Kordnoori S, Sabeti M, Shakoor MH, Moradi E. Deep multi-task learning structure for segmentation and classification of supratentorial brain tumors in MR images. Interdisciplinary Neurosurgery. 2024; 36: 101931.
• [14] Gurunathan A, Krishnan B. Detection and diagnosis of brain tumors using deep learning convolutional neural networks. Int J Imaging Syst Technol. 2021;31(3):1174–1184.
• [15] Eker AG, Duru N. Medikal görüntü işlemede derin öğrenme uygulamaları. Acta Infologica. 2021;5(2):459–474.
• [16] Dipu NM, Shohan SA, Salam KMA. Deep learning based brain tumor detection and classification. In: Proceedings of the 2021 International Conference on Intelligent Technologies (CONIT); 2021; 1–6. IEEE.
• [17] Jemimma TA, Vetharaj YJ. Watershed algorithm based DAPP features for brain tumor segmentation and classification. In: Proceedings of the 2018 International Conference on Smart Systems and Inventive Technology (ICSSIT); 2018; 155–158. IEEE.
• [18] Periasamy JK, Buvana S, Jeevitha P. Comparison of VGG-19 and RESNET-50 Algorithms in Brain Tumor Detection. In: Proceedings of the 2023 IEEE 8th International Conference for Convergence in Technology (I2CT); 2023; 1–5. IEEE.
• [19] Kumar T, Yadav PK, Yadav V. Detection of Brain Tumor using CNN. In: Proceedings of the 2022 4th International Conference on Inventive Research in Computing Applications (ICIRCA); 2022; 1121–1126. IEEE.
• [20] Prakash RM, Kumari R. Classification of MR brain images for detection of tumor with transfer learning from pre-trained CNN models. In: Proceedings of the 2019 International Conference on Wireless Communications Signal Processing and Networking (WiSPNET); 2019; 508–511. IEEE.
• [21] Khofiya SN, Fu’adah YN, Pratiwi NKC, Naufal RI, Pratama AD. Brain Tumor Classification Based On MRI Image Processing With Alexnet Architecture. In: Proceedings of the 2022 IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob); 2022; 1–6. IEEE.
• [22] Kumar M, Pilania U, Bhayana T, Thakur S. Utilizing YOLOv5x for the Detection and Classification of Brain Tumors. In: Proceedings of the 2024 2nd International Conference on Disruptive Technologies (ICDT); 2024; 1343–1348. IEEE.
• [23] Çınarer G, Emiroğlu BG. Classification of brain tumors by machine learning algorithms. In: Proceedings of the 2019 3rd International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT); 2019; 1–4. IEEE.
• [24] Aslan M. Derin Öğrenme Tabanlı Otomatik Beyin Tümör Tespiti. Fırat Üniv Mühendislik Bilimleri Derg. 2022;34(1):399–407.
• [25] Google. Google colab, 2017, https://colab.google/
• [26] Guido Van Rossum and Fred L Drake Jr. Python reference manual. Centrum voor Wiskunde en Informatica Amsterdam, 1995.
• [27] Keras Team. Keras: Deep Learning for humans, https://keras.io/
• [28] OpenCV. The OpenCV Reference Manual, 2.4.13.7 edition, April 2014.
• [29] Facebook. Pytorch, 2016, https://pytorch.org/
• [30] Brain Tumor — kaggle.com. https://www.kaggle.com/datasets/jakeshbohaju/brain-tumor/data, [Son erişim tarihi 06-01-2024].
• [31] Er MB. Akciğer Seslerinin Derin Öğrenme ile Sınıflandırılması. Gazi University Journal of Science Part C: Design and Technology. 2020;8(4):830–844.
• [32] Ali Ö. Automatic Detection of Epileptic Seizures from EEG Signals Using Artificial Intelligence Methods. Gazi University Journal of Science Part C: Design and Technology. 2024; 257–266.