A Deep Learning-Based Brain Tumor Classification System Using MRI Images from Bangladesh

AjajAnwarAnik ANWAR Anik,DurjoyKumerGhosh Kumer Ghosh,MahfuzurRahman  Rahman,MossaSamsurNaharSetu SamsurNahar Setu

ПРЕПРИНТ

Эта статья является препринтом и не была отрецензирована.
О результатах, изложенных в препринтах, не следует сообщать в СМИ как о проверенной информации.

A Deep Learning-Based Brain Tumor Classification System Using MRI Images from Bangladesh

A. A. Anik,

D. K. Ghosh,

M. Rahman,

M. S. Setu

2026-06-10

https://doi.org/10.24108/preprints-3115487

Abstract— This paper presents a deep learning-based system for brain tumor classification using Magnetic Resonance Imaging (MRI) images from a Bangladeshi dataset. Leveraging transfer learning with a DenseNet-121 architecture, the system is designed to accurately classify brain tumors into four distinct categories: Glioma, Meningioma, Pituitary, and Normal. The methodology involved comprehensive data preparation, including image resizing, augmentation, and careful splitting into training, validation, and test sets. The model was trained using the Adam optimizer and categorical cross-entropy loss, achieving a high test accuracy of 96.52% and robust performance across various metrics, including class-wise precision, recall, F1-score, specificity, and micro/macro-averaged ROC-AUC and PR-AUC scores. To enhance transparency and trustworthiness, explainability techniques such as Grad-CAM, Grad-CAM++, LayerCAM, and ScoreCAM were implemented to visualize the critical regions in MRI scans that influenced the model's predictions. These visualizations provided valuable insights into the model's decision-making process, confirming its focus on medically relevant features.

Ссылка для цитирования:

Anik A. A., Ghosh D. K., Rahman M., Setu M. S. 2026. A Deep Learning-Based Brain Tumor Classification System Using MRI Images from Bangladesh. PREPRINTS.RU. https://doi.org/10.24108/preprints-3115487

Список литературы

1. [1] Aamir, M., Rahman, Z., Bhatti, U.A. et al. An automated deep learning framework for brain tumor

2. classification using MRI imagery. Sci Rep 15, 17593 (2025). https://doi.org/10.1038/s41598-025-02209-2

3. [2] Ali, A., Li, X., Mashwani, W.K. et al. Multi-class brain tumor MRI segmentation and classification using

4. deep learning and machine learning approaches. Cancer Imaging 25, 131 (2025).

5. https://doi.org/10.1186/s40644-025-00953-2

6. [3] Dorfner, F.J., Patel, J.B., Kalpathy-Cramer, J. et al. A review of deep learning for brain tumor analysis in

7. MRI. npj Precis. Onc. 9, 2 (2025). https://doi.org/10.1038/s41698-024-00789-2

8. [4] Khan, M.A., Hussain, M.Z., Mehmood, S. et al. Transfer learning for accurate brain tumor classification

9. in MRI: a step forward in medical diagnostics. Discov Onc 16, 1040 (2025). https://doi.org/10.1007/s12672

10. 025-02671-4

11. [5] Yadav, R.K., Kumar, M. & Nandi, A. A Scalable Brain Tumor Diagnosis from Large-Scale MRI Datasets

12. Using CNN-ViT and Expert-Attention Fusions. Int J Comput Intell Syst 18, 254 (2025).

13. https://doi.org/10.1007/s44196-025-00981-7

14. [6] Aiya, A.J., Wani, N., Ramani, M. et al. Optimized deep learning for brain tumor detection: a hybrid

15. approach with attention mechanisms and clinical explainability. Sci Rep 15, 31386 (2025).

16. https://doi.org/10.1038/s41598-025-04591-3

17. [7] Komal Kumar Napa, Sangeetha Murugan, J.Senthil Murugan, A. Jayanthi,

18. BRAIN-META: A reproducible CNNâ€“vision transformer meta-ensemble pipeline for explainable brain

19. tumor classification,

20. MethodsX,Volume 16,2026,103769,ISSN 2215-0161,https://doi.org/10.1016/j.mex.2025.103769

21. Multi-Classification

22. [8] Gundogan, E. (2025). A Novel Hybrid Deep Learning Model Enhanced with Explainable AI for Brain Tumor Multi-Classification from MRI Images. Applied Sciences, 15(10), 5412. https://doi.org/10.3390/app15105412

23. [9] Afnaan, K., Arunbalaji, C.G., Singh, T. et al. Boosting brain tumor detection with an optimized ResNet

24. and explainability via Grad-CAM and LIME. Brain Inf. 12, 33 (2025). https://doi.org/10.1186/s40708-025

25. 00279-6

26. [10] Iftikhar, S., Anjum, N., Siddiqui, A.B. et al. Explainable CNN for brain tumor detection and

27. classification through XAI based key features identification. Brain Inf. 12, 10 (2025).

28. https://doi.org/10.1186/s40708-025-00257-y

29. [11] Adnan, K.M., Ghazal, T.M., Saleem, M. et al. Deep learning driven interpretable and informed decision

30. making model for brain tumour prediction using explainable AI. Sci Rep 15, 19223 (2025).

31. https://doi.org/10.1038/s41598-025-03358-0

32. [12] Md Shahriar Mannan, Prottoy; Chowdhury , Mahtab ; Rahman, Redwan ; Tamim , Azim Ullah ;

33. Rahman, Md Mizanur (2024), “PMRAM: Bangladeshi Brain Cancer - MRI Dataset ”, Mendeley Data, V1,

34. doi: 10.17632/m7w55sw88b.1

35. [13] Anand, V., Khajuria, A., Pachauri, R. K., & Gupta, V. (2026). Multi-class classification of brain tumors

36. using optimized CNN and transfer learning techniques. Scientific reports, 16(1), 4709.

37. https://doi.org/10.1038/s41598-025-34806-6

38. [14 ]Gisella, Luisa, Elena, Maquen-Niño., Gilberto, Carrión-Barco. (2023). Brain Tumor Classification Deep

39. Learning Model Using Neural Networks. International ARTICLE IN PRESS Journal

40. 10.3991/ijoe.v19i09.38819 of Online Engineering (ijoe), doi:

41. [15] A. Srivastava, A. Khare and A. Kushwaha, "Brain Tumor Classification using Deep Learning

42. Framework," 2023 International Conference on Intelligent Systems, Advanced Computing and

43. Communication (ISACC), Silchar, India, 2023, pp. 1-4, doi: 10.1109/ISACC56298.2023.10083818.

44. [16] Preeti, Jaidka., Sachin, Jain. (2023). Performance analysis of various deep learning techniques for brain

45. tumor classification.

46. [17] Q. Mastoi, S. Latif, S. Brohi, J. Ahmad, A. Alqhatani, M. S. Alshehri, A. Al Mazroa, and R. Ullah,

47. “Explainable AI in medical imaging: An interpretable and collaborative federated learning model for brain

48. tumor classification,” Frontiers in Oncology, vol. 15, Feb. 2025, doi: 10.3389/fonc.2025.1535478.

49. [18] P.Narayankar andV.P.Baligar, “Explainability of brain tumor classification based on region,” in 2024

50. International Conference on Emerging Technologies in Computer Science for Interdisciplinary Applications

51. (ICETCS), IEEE, Apr. 2024, 1–6, doi: 10.1109/icetcs61022.2024.10544289.

52. [19] Annadurai, Amrutha, Benoy Joseph, and Manas Ranjan Prusty. "Discrete Wavelet transform based

53. Multiscale Deep CNN cascaded LSTM model for the classification of Brain Tumor." (2023).

54. [20] Ke, L., Hu, G., Zhao, M., et al. (2026). Brain tumor classification from MRI images using a multi-scale

55. channel attention CNN integrated with SVM. Scientific Reports, 16, 6297.

56. [21] Anand, V., Khajuria, A., Pachauri, R. K., et al. (2026). Multi-class classification of brain tumors using

57. optimized CNN and transfer learning techniques. Scientific Reports, 16, 4709.

58. [22] Classification of Brain MRI Images using Deep Learning: The DeiT3 Model and the Use of Feature

59. Fusion Methods. (2026). Computers and Electronics in Medicine, 3(1), 77–85.

60. [23] Gomes, E. F., & Barbosa, R. S. (2026). Deep Learning Approaches for Brain Tumor Classification in

61. MRI Scans: An Analysis of Model Interpretability. Applied Sciences, 16(2), 831.

62. [24] Pant, K., Dutta Pramanik, P. K., & Zhao, Z. (2026). A Robust ConvNeXt-Based Framework for Efficient,

63. Generalizable, and Explainable Brain Tumor Classification on MRI. Bioengineering, 13(2), 157.

64. [25] L. Sánchez-Moreno, A. Perez-Peña, L. Duran-Lopez, and J. P. Dominguez- Morales, “Ensemble-based

65. convolutional neural networks for brain tumor classification in MRI: Enhancing accuracy and interpretability

66. using explainable AI,” Computers in Biology and Medicine, vol. 195, Art no. 110555, Sep. 2025, doi:

67. 10.1016/j.compbiomed.2025.11055 5

ПРЕПРИНТ

ИНФОРМАЦИЯ

Использование куки-файлов