INDEREX-NET: AN ATTENTION-GUIDED MULTI-STREAM CNN FRAMEWORK WITH ADAPTIVE FEATURE FUSION FOR BRAIN TUMOR CLASSIFICATION

Öz

Accurate classification of brain tumors from magnetic resonance imaging (MRI) data remains a challenging problem for computer-aided diagnosis systems due to inter-class similarity, intra-class variability, and heterogeneous tumor morphology. In this study, an attention-guided multi-stream deep learning architecture, termed IndereX-Net, is proposed for brain tumor classification from axial MRI images. The proposed model integrates three heterogeneous ImageNet-pretrained convolutional neural network backbones, namely ResNet50, DenseNet121, and InceptionV3, arranged in parallel streams to extract complementary multi-scale feature representations. Each stream is enhanced with a Convolutional Block Attention Module (CBAM), while Generalized Mean (GeM) pooling is employed for adaptive feature aggregation. In addition, a gated feature fusion mechanism is incorporated to dynamically regulate inter-stream interactions. The model is trained using a two-phase transfer learning strategy and evaluated on a publicly available brain tumor MRI dataset. Experimental results demonstrate that IndereX-Net outperforms individual backbone models, achieving an accuracy of 98.22% and a ROC-AUC of 0.9986, indicating that attention-guided multi-stream architectures provide an effective and reliable solution for computer-aided brain tumor diagnosis.

Anahtar Kelimeler

• Brain Tumor Classification
• Medical Image Analysis
• Attention-Based Deep Learning
• Multi-Stream Convolutional Neural Networks
• Adaptive Feature Fusion

INDEREX-NET: BEYİN TÜMÖRÜ SINIFLANDIRMASI İÇİN UYARLANABİLİR ÖZELLİK FÜZYONUNA SAHİP DİKKAT TABANLI ÇOK AKIŞLI CNN MİMARİSİ

Öz

Manyetik rezonans görüntüleme (MRI) verilerinden beyin tümörlerinin doğru sınıflandırılması; sınıflar arası benzerlik, sınıf içi değişkenlik ve tümör morfolojisindeki heterojenlik nedeniyle bilgisayar destekli tanı sistemleri açısından zorlu bir problemdir. Bu çalışmada, aksiyal MRI görüntülerinden beyin tümörü sınıflandırması için IndereX-Net adı verilen, dikkat mekanizması ile yönlendirilmiş çok akışlı bir derin öğrenme mimarisi önerilmektedir. Önerilen model; ResNet50, DenseNet121 ve InceptionV3 olmak üzere, ImageNet üzerinde önceden eğitilmiş üç heterojen evrişimsel sinir ağı omurgasını paralel akışlar halinde bütünleştirerek tamamlayıcı çok ölçekli özellik temsilleri elde etmektedir. Her bir akış Convolutional Block Attention Module (CBAM) ile güçlendirilmiş, uyarlanabilir özellik birleştirme için Generalized Mean (GeM) pooling kullanılmıştır. Ayrıca, akışlar arası etkileşimi dinamik olarak düzenleyen kapılı özellik füzyon mekanizması mimariye entegre edilmiştir. Model, iki aşamalı transfer öğrenme stratejisi ile eğitilmiş ve herkese açık bir beyin tümörü MRI veri seti üzerinde değerlendirilmiştir. Deneysel sonuçlar, IndereX-Net’in tekil omurga modellerine kıyasla daha üstün performans sergilediğini; %98,22 doğruluk ve 0,9986 ROC-AUC değerlerine ulaştığını göstererek, dikkat destekli çok akışlı mimarilerin bilgisayar destekli beyin tümörü tanısı için etkili ve güvenilir bir çözüm sunduğunu ortaya koymaktadır.

Anahtar Kelimeler

• Beyin Tümörü Sınıflandırması
• Tıbbi Görüntü Analizi
• Dikkat Tabanlı Derin Öğrenme
• Çok Akışlı Evrişimsel Sinir Ağları
• Uyarlanabilir Özellik Füzyonu

Kaynakça

1. Agadi, K., Dominari, A., Tebha, S.S., Mohammadi, A., Zahid, S., 2023. Neurosurgical management of cerebrospinal tumors in the era of artificial intelligence: a scoping review. Journal of Korean Neurosurgical Society, 66 (6), 632–641.
2. Ali, R.R., Yaacob, N.M., Alqaryouti, M.H., Sadeq, A.E., Doheir, M., Iqtait, M., Yaacob, S.S., 2025. Learning architecture for brain tumor classification based on deep convolutional neural network: Classic and ResNet50. Diagnostics, 15 (5), 624.
3. Alrikabi, J., Muhammad, S., 2024. Deep feature extraction with SVM classifier for brain tumor classification. Journal of Engineering and Physical Sciences, 14 (3).
4. Al-Saleh, A., Tejani, G.G., Mishra, S., Sharma, S.K., Mousavirad, S.J., 2025. A federated learning-based privacy-preserving image processing framework for brain tumor detection from CT scans. Scientific Reports, 15 (1), 23578.
5. Asif, S., Zhao, M., Tang, F., Zhu, Y., 2023. An enhanced deep learning method for multi-class brain tumor classification using deep transfer learning. Multimedia Tools and Applications, 82 (20), 31709–31736.
6. Bianchessi, T., Tampu, I.E., Blystad, I., Lundberg, P., Nyman, P., Eklund, A., Haj-Hosseini, N., 2023. Pediatric brain tumor type classification using deep learning on MR images from the Children’s Brain Tumor Network. medRxiv Preprint.
7. Cheng, D., Wang, Z., Li, J., 2023. Research on the application of CNN algorithm based on chaotic recursive diagonal model in medical image processing. Applied Mathematics and Nonlinear Sciences, 9 (1).
8. Chen, W., Sun, P., Zhang, Y., 2024. A brain tumor diagnosis approach based on deep separable convolutional networks and attention mechanisms. Applied Computational Engineering, 67 (1), 60–69.

9. Dong, R., Shiraiwa, A., Pawasut, A., Sreechun, K., Hayashi, T., 2024. Diagnosis of citrus greening using artificial intelligence: a Faster region-based convolutional neural network approach with convolution block attention module-integrated VGG-Net and ResNet models. Plants, 13 (12), 1631.
10. He, K., Zhang, X., Ren, S., Sun, J., 2016. Deep residual learning for image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA, 770– 778.
11. Hira, M.I.K., Hossain, M.S., Bithee, M.M.A., 2025. Brain tumor MRI dataset (glioma, meningioma, pituitary, no tumor). Mendeley Data, Version 1.
12. Hossain, A., Islam, M.T., Abdul Rahim, S.K., Rahman, M.A., Rahman, T., Arshad, H., Chowdhury, M.E., 2023. A lightweight deep learning-based microwave brain image network model for brain tumor classification using reconstructed microwave brain images. Biosensors, 13 (2), 238.
13. Hsu, S.P., Lin, M., Lin, C., Hsiao, T., Wang, Y., Sun, C., 2024. Brain tumor grading diagnosis using transfer learning based on optical coherence tomography. Biomedical Optics Express, 15 (4), 2343.
14. Huang, G., Liu, S., Van der Maaten, L., Weinberger, K.Q., 2018. CondenseNet: An efficient DenseNet using learned group convolutions. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, 2752–2761.
15. Ijaz, A., Raza, B., Kiran, I., Waheed, A., Raza, A., Shah, H., Aftan, S., 2023. Modality specific CBAM-VGGNet model for the classification of breast histopathology images via transfer learning. IEEE Access, 11, 15750–15762.
16. Janardhan, K., Thomas, C.B., 2025. Subclass classification of gliomas using MRI fusion technique. SN Computer Science, 6 (2), 1–15.
17. Khan, F., Ayoub, S., Gulzar, Y., Majid, M., Reegu, F., Mir, M., Elwasila, O., 2023. MRI-based effective ensemble frameworks for predicting human brain tumor. Journal of Imaging, 9 (8), 163.
18. Kochher, V., 2025. A comparative study of machine learning algorithms for brain tumor detection. Journal of Recent Innovations in Science and Technology, 1 (1), 1–9.
19. Kumar, K., Jyoti, K., 2024. Explainable AI in brain tumor diagnosis: A critical review of machine learning and deep learning techniques. Research Square Preprint.
20. Lu, Y., 2024. Multi-efficientnet ensemble-based brain tumor MRI images classification. Proceedings of the International Conference on Computer Graphics, Artificial Intelligence and Data Processing (ICCAID 2023).
21. Mastoi, Q.U.A., Latif, S., Brohi, S., Ahmad, J., Alqhatani, A., Alshehri, M.S., Ullah, R., 2025. Explainable artificial intelligence in medical imaging: an interpretable and collaborative federated learning model for brain tumor classification. Frontiers in Oncology, 15, 1535478.
22. Mohan, P., Ramkumar, G., 2024. Detection and classification of brain tumor using fine tuned MobileNet algorithm. Proceedings of the 3rd International Conference on Advanced Technologies (ICONAT), Goa, India, 1–5.
23. Murugesan, G., Nagendran, P., Natarajan, J., 2025. Advancing brain tumor diagnosis: deep siamese convolutional neural network as a superior model for MRI classification. Brain-X, 3 (2), e70028.
24. Pal, A., Seth, J., Tyagi, A., Haider, A., Mishra, A., 2024. An efficient brain tumor detection using ensemble learning. Emerging Advances in Technology and Practice.
25. Sumona, R.B., Biswas, J.P., Shafkat, A., Rahman, M.M., 2025. An integrated deep learning approach for enhancing brain tumor diagnosis. Healthcare Analytics.
26. Tariq, A., 2024. Transforming brain tumor diagnosis: vision transformers combined with ensemble techniques. Journal of Population Therapeutics and Clinical Pharmacology, 31, 1072–1084.
27. Zhang, J., Tan, X., Chen, W., Du, G., Fu, Q., Zhang, H., Jiang, H., 2023. EFF_D_SVM: a robust multi-type brain tumor classification system. Frontiers in Neuroscience, 17, 1269100.