Md. Tariqul Islam; Pintu Chandra Shill; Md Sadiq Iqbal

Comparative Analysis and Ensemble Optimization of CNN Architectures for MRI-Based Brain Tumor Diagnosis

PDF (1536KB), PP.167-188

Views: 0 Downloads: 0

Author(s)

Md. Tariqul Islam ^1,* Pintu Chandra Shill ¹ Md Sadiq Iqbal ²

1. Department of Mathematical and Physical Sciences, East West University, Dhaka, 1212, Bangladesh

2. Department of Computer Science and Engineering, Bangladesh University, Dhaka, 1207, Bangladesh

* Corresponding author.

DOI: https://doi.org/10.5815/ijigsp.2026.03.09

Received: 9 Mar. 2026 / Revised: 13 Apr. 2026 / Accepted: 2 May 2026 / Published: 8 Jun. 2026

Index Terms

MRI imaging, Machine Learning, Ensemble Learning, Brain Tumor, MobileNetV3

Abstract

Brain tumor detection and classification from MRI images is a challenging task. Early and accurate diagnosis are essential for selecting appropriate treatment plans and improving patient outcomes. Despite significant advances in deep learning for medical image recognition, comprehensive comparative analyses of brain tumor classification models, particularly regarding ensemble optimization, remain limited. This paper uses four state-of-the-art deep learning frameworks, namely EfficientNetB4, MobileNetV3, MobileNetV2, and EfficientNetB0, to classify brain MRI images into four categories: Glioma, Meningioma, Pituitary tumor, and Normal. It employs a two-phase transfer learning approach, followed by 5-fold cross-validation on 875 MRI images. A unified experimental framework is employed, incorporating a two-phase transfer learning approach, consistent preprocessing, and a rigorous evaluation protocol with 5-fold cross-validation and an independent test set to prevent data leakage. Both full and selective ensemble strategies are examined to improve the robustness and stability. The models are evaluated using accuracy, precision, recall, F-1 score, confusion matrices, and accuracy curves, and statistical validation using McNemar’s test. MobileNetV3 achieves the highest test accuracy of 98.76%, followed by EfficientNetB4 (97.89%) and EfficientNetB0 (93.48%). MobileNetV2 performs significantly worse, with an accuracy of less than 80%. The selective ensemble technique (which uses the best models) attains the highest accuracy of 92.97%, compared to the full ensemble (84.40%), which improves prediction robustness but does not surpass the best individual model in peak accuracy. Overall, it can be concluded that MobileNetV3 is the most suitable architecture for brain tumor classification, delivering high accuracy with minimal computational complexity. The selective ensemble approach also enhances performance, maintaining computational efficiency, emphasizing the importance of informed model selection in neuro-oncological image analysis and clinical decision-support systems.

Cite This Paper

Md Tariqul Islam, Pintu Chandra Shill, Md Sadiq Iqbal, "Comparative Analysis and Ensemble Optimization of CNN Architectures for MRI-Based Brain Tumor Diagnosis", International Journal of Image, Graphics and Signal Processing(IJIGSP), Vol.18, No.3, pp. 167-188, 2026. DOI:10.5815/ijigsp.2026.03.09

Reference

[1]Ostrom, Q. T., Cioffi, G., Waite, K., Kruchko, C., & Barnholtz-Sloan, J. S. (2023). CBTRUS statistical report: Primary brain and other central nervous system tumors diagnosed in the United States in 2016–2020. Neuro-Oncology, 25(Suppl 4), iv1–iv99. doi.org/10.1093/neuonc/noad149
[2]Claus, E. B., Bondy, M. L., Schildkraut, J. M., Wiemels, J. L., Wrensch, M., & Black, P. M. (2005). Epidemiology of intracranial meningioma. Neurosurgery, 57(6), 1088–1095. doi.org/10.1227/01.NEU.0000188281.91351.B9
[3]Stupp, R., Mason, W. P., van den Bent, M. J., Weller, M., Fisher, B., Taphoorn, M. J. B., … Mirimanoff, R. O. (2005). Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. New England Journal of Medicine, 352(10), 987– 996. doi.org/10.1056/NEJMoa043330
[4]Martucci M, et. al. Magnetic Resonance Imaging of Primary Adult Brain Tumors: State of the Art and Future Perspectives. Biomedicines. 2023 Jan 26;11(2):364. doi: 10.3390/biomedicines11020364.
[5]Louis, D. et. al. The 2016 World Health Organization classification of tumors of the central nervous system: A summary. Acta Neuropathologica, 131(6), 803–820. doi.org/10.1007/s00401-016-1545-1
[6]Law, M., Yang, S., Wang, H., Babb, J. S., Johnson, G., Cha, S., & Knopp, E. A. (2003). Glioma grading: Sensitivity, specificity, and predictive values of perfusion MR imaging and proton MR spectroscopic imaging compared with conventional MR imaging. American Journal of Neuroradiology, 24(10), 1989–1998.
[7]Osborn, A. G., Salzman, K. L., & Barkovich, A. J. (2017). Diagnostic imaging: Brain (3rd ed.)
[8]Hesamian, M. H., Jia, W., He, X., & Kennedy, P. (2023). Deep learning techniques for medical image segmentation: Achievements and challenges. Journal of Digital Imaging, 36(1), 1–21. doi.org/10.1007/s10278-022-00678-5
[9]Ahmad, I. S., Dai, J., Xie, Y., & Liang, X. (2025). Deep learning models for CT image classification: A comprehensive literature review. Quantitative Imaging in Medicine and Surgery, 15(1), 962–1011. doi.org/10.21037/qims-24-1400
[10]Ker, J., Wang, L., Rao, J., & Lim, T. (2022). Deep learning applications in medical image analysis. IEEE Access, 10, 70220–70245. https://doi.org/10.1109/ACCESS.2022.3189298
[11]Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009). ImageNet: A large-scale hierarchical image database. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 248–255). doi.org/10.1109/CVPR.2009.5206848
[12]Litjens, G., Kooi, T., Bejnordi, B. E., Setio, A. A. A., Ciompi, F., Ghafoorian, M., … Sánchez, C. I. (2017). A survey on deep learning in medical image analysis. Medical Image Analysis, 42, 60–88. doi.org/10.1016/j.media.2017.07.005
[13]Rundo, L., Han, C., & Maier, A. (2021). When convolutional neural networks meet medical imaging: A review of challenges and future trends. Electronics, 10(9), 1029. doi.org/10.3390/electronics10091029
[14]Hosseini, S. M. H., & Choi, J. Y. (2020). A comprehensive study on medical image data characteristics and deep learning challenges. Computers in Biology and Medicine, 120, 103736. doi.org/10.1016/j.compbiomed.2020.103736
[15]Tazin T, Sarker S, Gupta P, Ayaz F, Islam S, Monirujjaman Khan M, Bourouis S, Idris SA, Alshazly H. A Robust and Novel Approach for Brain Tumor Classification Using Convolutional Neural Network. Comput Intell Neurosci. 2021 Dec 21;2021:2392395. doi: 10.1155/2021/2392395.
[16]Shorten, C., & Khoshgoftaar, T. M. (2019). A survey on image data augmentation for deep learning. Journal of Big Data, 6, Article 60. https://doi.org/10.1186/s40537 019 0197 0
[17]Rastogi D, Johri P, Donelli M, Kumar L, Bindewari S, Raghav A, Khatri SK. Brain Tumor Detection and Prediction in MRI Images Utilizing a Fine-Tuned Transfer Learning Model Integrated Within Deep Learning Frameworks. Life (Basel). 2025Feb 20;15(3):327. doi: 10.3390/life15030327.
[18]Zahoor, M. M., Khan, S. H., Alahmadi, T. J., Alsahfi, T., Mazroa, A. S. A., Sakr, H. A., Alqahtani, S., Albanyan, A., & Alshemaimri, B. K. (2024). Brain Tumor MRI Classification Using a Novel Deep Residual and Regional CNN. Biomedicines, 12(7), 1395. https://doi.org/10.3390/biomedicines12071395
[19]S. Das, O. F. M. R. R. Aranya and N. N. Labiba, "Brain Tumor Classification Using Convolutional Neural Network," 2019 1st International Conference on Advances in Science, Engineering and Robotics Technology (ICASERT), Dhaka, Bangladesh, 2019, pp. 1-5, doi: 10.1109/ICASERT.2019.8934603.
[20]M. Tan, & Le, Q. V. (2019). EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks. In Proceedings of the 36th International Conference on Machine Learning (Vol. 97, pp. 6105–6114)
[21]Das, S., & Addya, R. K. (2025). Skin cancer detection and classification through medical image analysis using EfficientNet B0. NDT, 3(4), 23. doi.org/10.3390/ndt3040023
[22]Gencer, K. (2025). A comparative analysis of EfficientNetB0 and EfficientNetV2 variants for brain tumor classification using MRI images. International Journal of Innovative Engineering Applications, 9(1), 1–7. doi.org/10.46460/ijiea.1523782
[23]Zulfiqar, F., & Coauthors. (2023). Multi class classification of brain tumor types from MR images using EfficientNets. Biomedical Signal Processing and Control, 84, 104777. doi.org/10.1016/j.bspc.2023.104777
[24]Müller, D., Soto Rey, I., & Kramer, F. (2022). An analysis on ensemble learning optimized medical image classification with deep convolutional neural networks. IEEE Access, 10, 66467–66480. https://doi.org/10.1109/ACCESS.2022.3182399
[25]Geert Litjens et.al. A survey on deep learning in medical image analysis, Medical Image Analysis, Volume 42, 2017, Pages 60- 88, ISSN 1361-8415, doi.org/10.1016/j.media.2017.07.005.
[26]Das D, Biswas SK, Bandyopadhyay S. Detection of Diabetic Retinopathy using Convolutional Neural Networks for Feature Extraction and Classification (DRFEC). Multimed Tools Appl. 2022 Nov 29:1-59. doi: 10.1007/s11042-022-14165-4.
[27]Ghosh, S., & Chatterjee, A. (2023). Transfer ensemble learning based deep convolutional neural networks for diabetic retinopathy classification. arXiv. doi.org/10.48550/arXiv.2308.00525
[28]Supriyadi, M. R., Samah, A. B. A., Muliadi, J., Awang, R. A. R., Ismail, N. H., Abdul Majid, H., & Othman, M. S. B. (2025). A systematic literature review: Exploring the challenges of ensemble model for medical imaging. BMC Medical Imaging, 25, Article 128. https://doi.org/10.1186/s12880 025 01667 4
[29]Khan, A. A., Chaudhari, O., & Chandra, R. (2024). A review of ensemble learning and data augmentation models for class imbalanced problems: Combination, implementation and evaluation. Expert Systems with Applications, 244, 122778. doi.org/10.1016/j.eswa.2023.122778
[30]Shen, D., Wu, G., & Suk, H.-I. (2017). Deep learning in medical image analysis. Annual Review of Biomedical Engineering, 19, 221–248. doi.org/10.1146/annurev bioeng 071516 044442
[31]Alzubaidi, L., Duan, Y., Al Dujaili, A., Ibraheem, I. K., Alkenani, A. H., Santamaría, J., Fadhel, M. A., Al Shamma, O., Zhang, J., & Farhan, L. (2021). Deepening into the suitability of using pre trained models of ImageNet against a lightweight convolutional neural network in medical imaging: An experimental study. PeerJ Computer Science, 7, e715. doi.org/10.7717/peerj cs.715
[32]Rundo, L., Han, C., & Maier, A. (2021). When convolutional neural networks meet medical imaging: A review of challenges and future trends. Electronics, 10(9), 1029. https://doi.org/10.3390/electronics10091029
[33]Raghu, M., Zhang, C., Kleinberg, J. M., & Bengio, S. (2019). Transfusion: Understanding transfer learning for medical imaging. In Advances in Neural Information Processing Systems 32: Annual Conference on Neural Information Processing Systems 2019 (pp. 3342–3352).
[34]Wu Y, Chen B, Zeng A, Pan D, Wang R, Zhao S. Skin Cancer Classification With Deep Learning: A Systematic Review. Front Oncol. 2022 Jul 13;12:893972. doi: 10.3389/fonc.2022.893972.
[35]Zhou, Z., Rahman Siddiquee, M. M., Tajbakhsh, N., & Liang, J. (2019). UNet++: A nested U-Net architecture for medical image segmentation. Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support, 3–11. doi.org/10.1007/978-3-030-00889-5_1
[36]Menze, B. H., Jakab, A., Bauer, S., Kalpathy-Cramer, J., Farahani, K., Kirby, J., … van Leemput, K. (2015). The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS). IEEE Transactions on Medical Imaging, 34(10), 1993–2024. doi.org/10.1109/TMI.2014.2377694
[37]Valverde, S., Cabezas, M., Roura, E., González-Villà, S., Pareto, D., Vilanova, J. C., … Lladó, X. (2017). Improving automated multiple sclerosis lesion segmentation with a cascaded 3D convolutional neural network approach. NeuroImage, 155, 159–168. doi.org/10.1016/j.neuroimage.2017.04.038
[38]Jun Cheng. Brain Tumor Dataset [DS/OL]. V1. Science Data Bank, 2022 Accessed:2025-12-01. CSTR: 31253.11.sciencedb.06290.
[39]Figshare Brain Tumor MRI Dataset - DOI: https://doi.org/10.6084/m9.figshare.14778750
[40]Integrated Brain Tumor MRI Dataset - A combined dataset created by merging 26 and 27 for the purpose of model development and evaluation in this study.
[41]Rastogi, D., et al. (2025). Brain Tumor Detection and Prediction in MRI Images Utilizing a Fine-Tuned Transfer Learning Model Integrated Within Deep Learning Frameworks. Life, 15(3), 327. doi.org/10.3390/life15030327
[42]Khan, M.A., et al. (2025). Transfer learning for accurate brain tumor classification in MRI: a step forward in medical diagnostics. Discov Onc 16, 1040 doi.org/10.1007/s12672-025-02671-4
[43]Babu Vimala B, Srinivasan S, Mathivanan SK, Mahalakshmi, Jayagopal P, Dalu GT. Detection and classification of brain tumor using hybrid deep learning models. Sci Rep. 2023 Dec 27;13(1):23029. doi: 10.1038/s41598-023-50505-6.
[44]Farandi, M.N., Muda, A.K., Winarno, S., & Basiron, H. (2025). Comparative Study of Deep Learning Models for MRI-based Brain Tumor Classification. Journal of Future Artificial Intelligence and Technologies.
[45]Rahim Khan, et. al.(2025). High-precision brain tumor classification from MRI images using an advanced hybrid deep learning method with minimal radiation exposure, Journal of Radiation Research and Applied Sciences, Volume 18, Issue 4,2025,101858, ISSN 1687-8507, doi.org/10.1016/j.jrras.2025.101858.

International Journal of Image, Graphics and Signal Processing (IJIGSP)