Authors : Ameer Hamza; Yihong Zhang; Md Saifur Rahman; Shijun Sun; Yaoyao Ran; Ali Sajid; Muhammad Abubakr

Volume/Issue : Volume 11 - 2026, Issue 5 - May

Google Scholar : https://tinyurl.com/4fh93kv7

Scribd : https://tinyurl.com/466sj3kx

DOI : https://doi.org/10.38124/ijisrt/26May529

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Abstract : Accurate segmentation of brain tumour sub-regions from multi-modal MRI remains clinically challenging, as conventional three-dimensional U-Net architectures are hampered by vanishing gradients, indiscriminate skip-connection propagation, and insufficient multi-scale supervision, collectively limiting robust delineation of heterogeneous tumour components on the BraTS 2020 benchmark. An Enhanced Three-Dimensional Attention U-Net with Deep Supervision is proposed, trained on 128³ isotropic volumes from three MRI modalities (FLAIR, T1ce, T2) using stochastic multitransform augmentation and a composite class-weighted Dice–focal–boundary loss to jointly address class imbalance and imprecise enhancing tumour delineation. The architecture incorporates a fully residual encoder–decoder backbone with graduated spatial dropout (0.10–0.30), soft spatial attention gates at every skip connection to suppress background activation, and a resolution-aware deep supervision scheme weights with morphological post-processing to enforce anatomical plausibility. The proposed method achieves Dice scores of 0.817, 0.811, and 0.846 for enhancing tumour, tumour core, and whole tumour respectively, with HD95 values of 2.95, 3.24, and 3.97 mm, demonstrating superior boundary precision over nnU-Net, Swin UNETR, TransUNet, H2NF-Net, and ACU-Net, confirming the clinical viability of the integrated framework for precise multi-class brain tumour segmentation.

Keywords : Brain Tumour Segmentation, Multimodal MRI, Attention U-Net, Residual Learning, Deep Supervision, Brats 2020, Class-Weighted Loss, Tumour Delineation.

References :

1. Dorfner, F.J., et al., A review of deep learning for brain tumor analysis in MRI. npj Precision Oncology, 2025. 9(1): p. 2.
2. Bonato, B., L. Nanni, and A. Bertoldo, Advancing Precision: A Comprehensive Review of MRI Segmentation Datasets from BraTS Challenges (2012-2025). Sensors (Basel), 2025. 25(6).
3. Lin, S.Y. and C.L. Lin, Brain tumor segmentation using U-Net in conjunction with EfficientNet. PeerJ Comput Sci, 2024. 10: p. e1754.
4. R, P., J.P.P. M, and N. J S, Brain tumor segmentation using multi-scale attention U-Net with EfficientNetB4 encoder for enhanced MRI analysis. Scientific Reports, 2025. 15(1): p. 9914.
5. Verma, A. and A.K. Yadav, Brain tumor segmentation with deep learning: Current approaches and future perspectives. Journal of Neuroscience Methods, 2025. 418: p. 110424.
6. Yeafi, A., M. Islam, and M.S.U. Yusuf, A deep learning framework for 3D brain tumor segmentation and survival prediction. Healthcare Analytics, 2025. 8: p. 100418.
7. Feng, X., et al., Brain Tumor Segmentation Using an Ensemble of 3D U-Nets and Overall Survival Prediction Using Radiomic Features. Frontiers in Computational Neuroscience, 2020. Volume 14 - 2020.
8. Raza, R., et al., dResU-Net: 3D deep residual U-Net based brain tumor segmentation from multimodal MRI. Biomedical Signal Processing and Control, 2023. 79: p. 103861.
9. Lin, W.W., et al., A novel 2-phase residual U-net algorithm combined with optimal mass transportation for 3D brain tumor detection and segmentation. Sci Rep, 2022. 12(1): p. 6452.
10. Naqvi, N.Z. and K.R. Seeja, An Attention-Based Residual U-Net for Tumour Segmentation Using Multi-Modal MRI Brain Images. IEEE Access, 2025. 13: p. 10240–10251.
11. Maqsood, R., et al., Optimal Res-UNET architecture with deep supervision for tumor segmentation. Front Med (Lausanne), 2025. 12: p. 1593016.
12. Abrar, M., et al., Enhancing brain tumor segmentation using attention based convolutional UNet on MRI images. Scientific Reports, 2025. 15(1): p. 36603.
13. Huang, J., et al., Brain tumor segmentation using deep learning: high performance with minimized MRI data. Frontiers in Radiology, 2025. Volume 5 - 2025.
14. Aumente-Maestro, C., et al., BTS U-Net: A data-driven approach to brain tumor segmentation through deep learning. Biomedical Signal Processing and Control, 2025. 104: p. 107490.
15. Chen, W., et al., MAUNet: a mixed attention U-net with spatial multi-dimensional convolution and contextual feature calibration for 3D brain tumor segmentation in multimodal MRI. Frontiers in Neuroscience, 2025. Volume 19 - 2025.
16. Talukder, M.A., M. Tabassum, and M. Khalid, Self-attention U-Net (SAU-Net): An attention-driven U-Net framework for precise brain tumor segmentation using multimodal magnetic resonance imaging. Digit Health, 2026. 12: p. 20552076261426312.
17. Mishra, S., et al., Data-Driven Deep Supervision for Medical Image Segmentation. IEEE Trans Med Imaging, 2022. 41(6): p. 1560–1574.
18. Fu, Z., et al., Deep supervision feature refinement attention network for medical image segmentation. Engineering Applications of Artificial Intelligence, 2023. 125: p. 106666.
19. Turečková, A., et al., Improving CT Image Tumor Segmentation Through Deep Supervision and Attentional Gates. Frontiers in Robotics and AI, 2020. Volume 7 - 2020.
20. Maqsood, R., et al., Optimal Res-UNET architecture with deep supervision for tumor segmentation. Frontiers in Medicine, 2025. Volume 12 - 2025.
21. Gao, Y., et al., Medical Image Segmentation: A Comprehensive Review of Deep Learning-Based Methods. Tomography, 2025. 11(5).
22. Ma, J., et al., Loss odyssey in medical image segmentation. Medical Image Analysis, 2021. 71: p. 102035.
23. Ma, J., et al., Loss odyssey in medical image segmentation. Med Image Anal, 2021. 71: p. 102035.
24. Li, B., et al., Region-related focal loss for 3D brain tumor MRI segmentation. Med Phys, 2023. 50(7): p. 4325–4339.
25. Yeung, M., et al., Unified Focal loss: Generalising Dice and cross entropy-based losses to handle class imbalanced medical image segmentation. Comput Med Imaging Graph, 2022. 95: p. 102026.
26. Visalakshi, G. and L. Mohan, A novel sub-differentiable hausdorff loss combined with BCE for MRI brain tumor segmentation using UNet variants. Scientific Reports, 2025. 15(1): p. 45136.
27. Zhou, T., S. Ruan, and B. Lei, BUFNet: Boundary-aware and uncertainty-driven multi-modal fusion network for MR brain tumor segmentation. Medical Image Analysis, 2026. 107: p. 103855.
28. Yang, L., et al., MUNet: a novel framework for accurate brain tumor segmentation combining UNet and mamba networks. Frontiers in Computational Neuroscience, 2025. Volume 19 - 2025.
29. Kofler, F., et al., BraTS Toolkit: Translating BraTS Brain Tumor Segmentation Algorithms Into Clinical and Scientific Practice. Frontiers in Neuroscience, 2020. Volume 14 - 2020.
30. Jiang, Z., et al., ENHANCING GENERALIZABILITY IN BRAIN TUMOR SEGMENTATION: MODEL ENSEMBLE WITH ADAPTIVE POST-PROCESSING. Proc IEEE Int Symp Biomed Imaging, 2024. 2024.
31. Banerjee, S. and S. Mitra, Novel Volumetric Sub-region Segmentation in Brain Tumors. Frontiers in Computational Neuroscience, 2020. Volume 14 - 2020.
32. Hardani, D.N.K., H.A. Nugroho, and I. Ardiyanto. An Automatic Brain Tumor Segmentation Using 3D Residual U-Net. in 2022 6th International Conference on Information Technology, Information Systems and Electrical Engineering (ICITISEE). 2022.
33. Schlemper, J., et al., Attention gated networks: Learning to leverage salient regions in medical images. Medical Image Analysis, 2019. 53: p. 197–207.
34. Alshomrani, S., M. Arif, and M.A. Al Ghamdi, SAA-UNet: Spatial Attention and Attention Gate UNet for COVID-19 Pneumonia Segmentation from Computed Tomography. Diagnostics (Basel), 2023. 13(9).
35. He, X., et al., Deep Convolutional Neural Network With a Multi-Scale Attention Feature Fusion Module for Segmentation of Multimodal Brain Tumor. Frontiers in Neuroscience, 2021. Volume 15 - 2021.
36. Rehman, A., et al., Selective Deeply Supervised Multi-Scale Attention Network for Brain Tumor Segmentation. Sensors (Basel), 2023. 23(4).
37. Yang, J., et al., MSDS-UNet: A multi-scale deeply supervised 3D U-Net for automatic segmentation of lung tumor in CT. Computerized Medical Imaging and Graphics, 2021. 92: p. 101957.
38. Park, M., et al., ES-UNet: efficient 3D medical image segmentation with enhanced skip connections in 3D UNet. BMC Med Imaging, 2025. 25(1): p. 327.
39. Yuan, D., et al., μ-Net: Medical image segmentation using efficient and effective deep supervision. Computers in Biology and Medicine, 2023. 160: p. 106963.
40. Huang, D., et al., Learning rich features with hybrid loss for brain tumor segmentation. BMC Med Inform Decis Mak, 2021. 21(Suppl 2): p. 63.
41. Montazerolghaem, M., et al., U-Net Architecture for Prostate Segmentation: The Impact of Loss Function on System Performance. Bioengineering (Basel), 2023. 10(4).
42. Liu, Z., et al., Innovative multi-class segmentation for brain tumor MRI using noise diffusion probability models and enhancing tumor boundary recognition. Scientific Reports, 2024. 14(1): p. 29576.
43. Isensee, F., et al., Automated design of deep learning methods for biomedical image segmentation. arXiv preprint arXiv:1904.08128, 2019.
44. Abrar, M., et al., Enhancing brain tumor segmentation using attention based convolutional UNet on MRI images. Sci Rep, 2025. 15(1): p. 36603.
45. Chen, J., et al., TransUNet: Rethinking the U-Net architecture design for medical image segmentation through the lens of transformers. Medical Image Analysis, 2024. 97: p. 103280.
46. Hatamizadeh, A., et al. Swin unetr: Swin transformers for semantic segmentation of brain tumors in mri images. in International MICCAI brainlesion workshop. 2021. Springer.
47. Wu, X., et al., FCFDiff-Net: full-conditional feature diffusion embedded network for 3D brain tumor segmentation. Quant Imaging Med Surg, 2025. 15(5): p. 4217–4234.
48. Chen, W., et al., MAUNet: a mixed attention U-net with spatial multi-dimensional convolution and contextual feature calibration for 3D brain tumor segmentation in multimodal MRI. Frontiers in Neuroscience, 2025. Volume 19 - 2025.