Segmentation of Medical MRI Images Using Nested U-Net with Attention Mechanism and Fuzzy Pooling

Basheer, Noor M.; Al-Saegh, Ali

doi:10.33899/arej.2025.159398.1409

	Segmentation of Medical MRI Images Using Nested U-Net with Attention Mechanism and Fuzzy Pooling
Al-Rafidain Engineering Journal (AREJ)
Volume 30, Issue 2, September 2025, Pages 104-115 PDF (797.44 K)
Document Type: Research Paper
DOI: 10.33899/arej.2025.159398.1409
Authors
Noor M. Basheer; Ali Al-Saegh^*
Computer Engineering Department, College of Engineering, University of Mosul, Mosul, Iraq
Abstract
Segmentation of medical images is a crucial step in the diagnosis and treatment of various diseases, particularly when analyzing magnetic resonance imaging (MRI) scans. Despite significant progress, accurate segmentation remains challenging due to the complexity of anatomical structures, ambiguous boundaries, and variations in image quality. This paper proposes an enhanced Nested U-Net structure that incorporates an attention mechanism and fuzzy pooling to improve retail performance. Nested U-Net benefits from extensive skip connections to capture multi-level contextual information, while the attention mechanism enhances the model's ability to focus on relevant features and noise suppression. In addition, we replace the traditional extreme pooling layers with fuzzy pooling, which enables the network to handle spatial ambiguity more effectively and produce more accurate and refined retail maps. Experimental results on publicly available MRI datasets show that the proposed model achieves remarkable improvements, with a 4.97% increase in the dice coefficient, a 2.76% improvement in accuracy, and a 4.25% increase in recall compared to enhanced U-Net structures. Furthermore, significant improvements have been observed in Hausdorff's distance measures.
Keywords
Attention Mechanism; Deep Learning in Medical Imaging; Fuzzy Pooling; Nested U-Net; MRI Image Segmentation

References
[1] Y. M. A. Mohammed, S. El Garouani, and I. Jellouli, “A survey of methods for brain tumor segmentation-based MRI images,” J. Comput. Des. Eng., vol. 10, no. 1, pp. 266–293, 2023, doi: 10.1093/jcde/qwac141. [2] A. Kumar, “Study and analysis of different segmentation methods for brain tumor MRI application,” Multimed. Tools Appl., vol. 82, no. 5, pp. 7117–7139, 2023, doi: 10.1007/s11042-022-13636-y. [3] N. Abdullah and S. Al-Kazzaz, “Evaluation of Physiotherapy Exercise by Motion Capturing Based on Artificial Intelligence: A Review,” Al-Rafidain Engineering Journal (AREJ), 28, no. 2, pp. 237–251, 2023, doi: 10.33899/rengj.2023.138562.1235. [4] R. Yousef, G. Gupta, N. Yousef, and M. Khari, A holistic overview of deep learning approach in medical imaging, vol. 28, no. 3. Springer Berlin Heidelberg, 2022. doi: 10.1007/s00530-021-00884-5. [5] H. Cui, H. Pan, and K. Zhang, “SCU-Net++: A Nested U-Net Based on Sharpening Filter and Channel Attention Mechanism,” Wirel. Commun. Mob. Comput., vol. 2022, 2022, doi: 10.1155/2022/2848365. [6] R. A. Moyassar and M. A. M. Abdullah, “Advancing EMG Finger Movement Classification with Feature Extraction and Machine Learning,” Al-Rafidain Eng. J., vol. 30, no. 1, pp. 150–163, 2025, doi: 10.33899/arej.2024.152539.1368. [7] A. Al-Saegh, S. A. Dawwd, and J. M. Abdul-Jabbar, “Towards Efficient Motor Imagery EEG-based BCI Systems using DCNN,” 2024 Arab ICT Conf. AICTC 2024, pp. 59–66, 2024, doi: 10.1109/AICTC58357.2024.10735028. [8] H. Peng and S. Yu, “A Systematic IoU-Related Method: Beyond Simplified Regression for Better Localization,” IEEE Trans. Image Process., vol. 30, no. i, pp. 5032–5044, 2021, doi: 10.1109/TIP.2021.3077144. [9] Z. Fu, H. Chen, M. He, and L. Liu, “An Enhanced U-Network by Combining PPM and CBAM for Medical Image Segmentation,” IEEE Access, vol. 12, no. June, pp. 107098–107112, 2024, doi: 10.1109/ACCESS.2024.3426518. [10] S. R. Gunasekara, H. N. T. K. Kaldera, and M. B. Dissanayake, “A systematic approach for MRI brain tumor localization and segmentation using deep learning and active contouring,” J. Healthc. Eng., vol. 2021, pp. 1–13, 2021, doi: 10.1155/2021/6695108. [11] M. A. Ottom, H. A. Rahman, and I. D. Dinov, “Znet: Deep learning approach for 2D MRI brain tumor segmentation,” IEEE J. Transl. Eng. Health Med., vol. 10, no. 1, pp. 1–8, 2022, doi: 10.1109/JTEHM.2022.3176737. [12] X. Guan et al., “3D AGSE-VNet: an automatic brain tumor MRI data segmentation framework,” BMC Med. Imaging, vol. 22, no. 1, pp. 1–18, 2022, doi: 10.1186/s12880-021-00728-8. [13] A. Al-Saegh, A. Daood, and M. H. Ismail, “Dual Optimization of Deep CNN for Motor Imagery EEG Tasks Classification,” Diyala J. Eng. Sci., vol. 17, no. 4, pp. 75–91, 2024, doi: 10.24237/djes.2024.17405. [14] B. Chen, Y. Liu, Z. Zhang, G. Lu, and A. W. K. Kong, “TransAttUnet: Multi-Level Attention-Guided U-Net with Transformer for Medical Image Segmentation,” IEEE Trans. Emerg. Top. Comput. Intell., vol. 8, no. 1, pp. 55–68, 2024, doi: 10.1109/TETCI.2023.3309626. [15] S. Sinha, M. Bhatt, A. Anand, A. S. Areeckal, and V. Aparna, “EnigmaNet: A Novel Attention-Enhanced Segmentation Framework for Ischemic Stroke Lesion Detection in Brain MRI,” IEEE Access, vol. 12, no. June, pp. 91480–91498, 2024, doi: 10.1109/ACCESS.2024.3422025. [16] G. Han et al., “Improved U-Net based insulator image segmentation method based on attention mechanism,” Energy Reports, vol. 7, pp. 210–217, 2021, doi: 10.1016/j.egyr.2021.10.037. [17] L. Shen, Q. Wang, W. Wang, R. Wang, Y. Yu, and C. Zhang, “Deep attention enhanced networks for medical image segmentation,” vol. 13167, no. Rsmip, p. 126, 2024, doi: 10.1117/12.3029812. [18] Y. Zhou, Q. Kong, Y. Zhu, and Z. Su, “MCFA-UNet: Multiscale Cascaded Feature Attention U-Net for Liver Segmentation,” Irbm, vol. 44, no. 4, p. 100789, 2023, doi: 10.1016/j.irbm.2023.100789. [19] M. T. Rostami and A. Motamedi, “IRA-Unet: Inception Residual Attention U-Net in adversarial network for cardiac MRI segmentation,” TechRxiv, preprint, pp. 1–10, Aug. 2023, doi: 10.36227/techrxiv.23896641. [20] S. Kuang et al., “MSCDA: Multi-level semantic-guided contrast improves unsupervised domain adaptation for breast MRI segmentation in small datasets,” Neural Networks, vol. 165, pp. 119–134, 2023, doi: 10.1016/j.neunet.2023.05.014. [21] F. Zhang et al., “Deep learning based segmentation of brain tissue from diffusion MRI,” Neuroimage, vol. 233, no. July 2020, p. 117934, 2021, doi: 10.1016/j.neuroimage.2021.117934. [22] Y. Gu, Z. Piao, and S. J. Yoo, “STHarDNet: Swin Transformer with HarDNet for MRI Segmentation,” Appl. Sci., vol. 12, no. 1, 2022, doi: 10.3390/app12010468. [23] C. Zhang, A. Achuthan, and G. M. S. Himel, “State-of-the-Art and Challenges in Pancreatic CT Segmentation: A Systematic Review of U-Net and Its Variants,” IEEE Access, vol. 12, no. June, pp. 78726–78742, 2024, doi: 10.1109/ACCESS.2024.3392595. [24] H. Wang, S. Qiu, B. Zhang, and L. Xiao, “Multilevel Attention Unet Segmentation Algorithm for Lung Cancer Based on CT Images,” Comput. Mater. Contin., vol. 78, no. 2, pp. 1569–1589, 2024, doi: 10.32604/cmc.2023.046821. [25] K. Wisaeng, “U-Net++DSM: Improved U-Net++ for Brain Tumor Segmentation With Deep Supervision Mechanism,” IEEE Access, vol. 11, no. October, pp. 132268–132285, 2023, doi: 10.1109/ACCESS.2023.3331025. [26] T. Zhou, X. Ye, H. Lu, X. Zheng, S. Qiu, and Y. Liu, “Dense convolutional network and its application in medical image analysis,” Biomed. Res. Int., vol. 2022, pp. 1–11, 2022, doi: 10.1155/2022/2384830. [27] Z. Mahmoud and M. A. M. Abdullah, “Comparison of Deep Learning and Machine Learning Techniques for Automated Diagnosis of Acute Lymphoblastic Leukemia,” Al-Rafidain Eng. J., vol. 30, no. 1, pp. 175–185, 2025, doi: 10.33899/arej.2024.152478.1366. [28]Z. Li, H. Zhang, Z. Li, and Z. Ren, “Residual-attention UNet++: A nested residual-attention U-Net for medical image segmentation,” Appl. Sci., vol. 12, no. 14, pp. 1–14, 2022, doi: 10.3390/app12147149. [29] Z. Al-Mokhtar, F. Ibraheem, and H. Al-Layla, “A Review of Digital Image Fusion and its Application,” Al-Rafidain Eng. J., vol. 26, no. 2, pp. 309–322, 2021, doi: 10.33899/rengj.2021.127928.1055. [30] M. Cui, K. Li, J. Chen, and W. Yu, “CM-Unet: A Novel Remote Sensing Image Segmentation Method Based on Improved U-Net,” IEEE Access, vol. 11, no. June, pp. 56994–57005, 2023, doi: 10.1109/ACCESS.2023.3282778. [31] A. Al-Saegh, “Identifying a Suitable Signal Processing Technique for MI EEG Data,” Tikrit J. Eng. Sci., vol. 30, no. 3, pp. 140–147, 2023, doi: 10.25130/tjes.30.3.14. [32] N. S. Punn and S. Agarwal, “Modality specific U-Net variants for biomedical image segmentation: A survey,” Artif. Intell. Rev., vol. 55, no. 7, pp. 5845–5889, 2022, doi: 10.1007/s10462-022-10152-1. [33]M. Ali, S. O. Gilani, A. Waris, K. Zafar, and M. Jamil, “Brain Tumour Image Segmentation Using Deep Networks,” IEEE Access, vol. 8, pp. 153589–153598, 2020, doi: 10.1109/ACCESS.2020.3018160. [34] O. Oktay et al., “Attention U-Net: Learning where to look for the pancreas,” arXiv preprint, 2018. [Online]. Available: http://arxiv.org/abs/1804.03999 [35] A. Dosovitskiy et al., “An image is worth 16×16 words: Transformers for image recognition at scale,” in Proc. Int. Conf. Learn. Represent. (ICLR), 2021. [Online]. Available: http://arxiv.org/abs/2010.11929 [36] J. Chen et al., “TransUNet: Transformers make strong encoders for medical image segmentation,” arXiv preprint, pp. 1–13, 2021. [Online]. Available: http://arxiv.org/abs/2102.04306 [37] H. Cao et al., “Swin-Unet: Unet-Like Pure Transformer for Medical Image Segmentation,” Lect. Notes Comput. Sci. (including Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinformatics), vol. 13803 LNCS, pp. 205–218, 2023, doi: 10.1007/978-3-031-25066-8_9. [38] G. Xu, X. Zhang, X. He, and X. Wu, “LeViT-UNet: Make Faster Encoders with Transformer for Medical Image Segmentation,” Lect. Notes Comput. Sci. (including Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinformatics), vol. 14432 LNCS, pp. 42–53, 2024, doi: 10.1007/978-981-99-8543-2_4. [39] A. Tragakis, C. Kaul, R. Murray-Smith, and D. Husmeier, “The Fully Convolutional Transformer for Medical Image Segmentation,” Proc. - 2023 IEEE Winter Conf. Appl. Comput. Vision, WACV 2023, pp. 3649–3658, 2023, doi: 10.1109/WACV56688.2023.00365. [40] H.-Y. Zhou, J. Guo, Y. Zhang, L. Yu, L. Wang, and Y. Yu, “nnFormer: Interleaved Transformer for volumetric segmentation,” arXiv preprint, pp. 1–10, 2021. [Online]. Available: http://arxiv.org/abs/2109.03201 [41] M. Z. Alom, C. Yakopcic, M. Hasan, T. M. Taha, and V. K. Asari, “Recurrent residual U-Net for medical image segmentation,” Journal of Medical Imaging, vol. 6, no. 1, pp. 014006-1–014006-13, 2019, doi: 10.1117/1.jmi.6.1.014006. [42] C. Li et al., “ANU-Net: Attention-based nested U-Net to exploit full resolution features for medical image segmentation,” Comput. Graph., vol. 90, pp. 11–20, 2020, doi: 10.1016/j.cag.2020.05.003.
Statistics Article View: 346 PDF Download: 171