Segmentation Methods for MS Lesions in MRI Images Using U-Net-Based Architectures

Document Type : Original Article

Authors

1 PhD Student, Computer Engineering Department, Yazd University, Yazd, Iran

2 Professor, Computer Engineering Department, Yazd University, Yazd, Iran

3 Associate Professor, Computer Engineering Department, Yazd University, Yazd, Iran

4 Associate Professor, Shahid Sadoughi University of Medical Sciences, Yazd, Iran

Abstract

MS is one of the most common autoimmune disorders of the central nervous system. It is characterized by the destruction of the myelin sheath of nerve cells and the formation of lesions in brain and spinal cord tissues. Accurate detection of lesions in MRI images plays a crucial role in diagnosing the disease, predicting patient recovery, and selecting appropriate treatment methods. In this article, with a review and analytical approach, the performance of three widely used deep learning architectures including U-Net, ResUNet, and AttentionUNet in the segmentation of MS lesions is examined and compared. After introducing commonly used datasets and evaluation metrics in this field, results from 34 reliable studies were analyzed using the Dice coefficient. On average, U-Net, ResUNet, and AttentionUNet architectures achieved Dice scores of 0.6562, 0.6937, and 0.7435 respectively in MS lesion segmentation in the reviewed studies. The results show that AttentionUNet performs better than other methods. Also, the strengths and weaknesses of each method are analyzed and future research directions in this area are proposed.

Keywords

Main Subjects

References

[1]     M. Rakić, S. Vercruyssen, S. Eyndhoven and E. Rosa, "Icobrain ms 5.1: combining unsupervised and supervised approaches for improving the detection of multiple sclerosis lesions," NeuroImage. Clinical, vol. 31, p. 102707, 2021
[2]     N. Aslam, I. Khan, A. Bashamakh and F. Alghool, "Multiple sclerosis diagnosis using machine learning and deep learning: challenges and opportunities," Sensors, vol. 22, no. 20, p. 7856, 2022.
[3]     Definition of MS. (2025, June. 7). [Online]. Available: https: //dr-moradi.com.
[4]     A. Alijamaat, A. NikravanShalmani and P. Bayat, "Multiple sclerosis identification in brain MRI images using wavelet convolutional neural networks," International Journal of Imaging Systems and Technology, 2020.
[5]     NeuroMS. (2023, May. 11). [Online]. Available: https://royancell.ir/NeuroMSCellT.html.
[6]     M. Moghadasi and G. Fazekas, "Segmentation of MRI images to detect multiple sclerosis using non-parametric, non-uniform intensity normalization and support vector machine methods," Infocommunications journal, vol. 13, no. 1, pp. 68–74, 2021.
[7]     NeuroImaging Tools and Resources Collaboratory. (Accessed 16 May 2023) [Online]. Available: https://www.nitrc.org/projects/msseg/.
[8]     M. Styner, J. Lee, B. Chin, M. Commowick and C. Olivier, "3D segmentation in the clinic: a grand challenge II: MS lesion segmentation," Researchgate, 2008.
[9]     A. Carass , R. Snehashis, J. Amod and C. Jennifer,"Longitudinal multiple sclerosis lesion segmentation data resource," Data in Brief, vol. 12, pp. 346–350, 2017
[10] O. Commowick ,K. Michaël, C. Romain, "Multiple sclerosis lesions segmentation from multiple experts: The MICCAI 2016 challenge dataset," NeuroImage, vol. 244, p. 118589, 2021
[11] "shanoir,"Irisa.fr, (2023, May. 11) [Online]. Available: https://shanoir.irisa.fr/shanoir-ng/welcome.
[12] A. M Muslim, "Brain MRI dataset of multiple sclerosis with consensus manual lesion segmentation and patient meta information," data.mendeley.com, vol. 1, 2022.
[13] Center for Biomedical Image Computing & Analytics, (Accessed 16 May 2023). [Online]. Available: http://braintumorsegmentation.org/.
[14] L. Baskaran , S. Al’Aref, G. Maliakal, "Automatic segmentation of multiple cardiovascular structures from cardiac computed tomography angiography images using deep learning," plosone, vol. 15, no. 5, 2020.
[15] Dice coefficient, IOU. (2025, June. 7). [Online]. Available: https://karan-jakhar.medium.com.
[16] O. Ronneberger, P. Fischer, and T. Brox, "U-Net: convolutional networks for biomedical image segmentation," Lecture Notes in Computer Science, vol. 9351, pp. 234–241, 2015
[17] R. Nair, T. Babu, G. Ramasamy, T. Singh and X. Yuan, "Advanced U-Net++ architecture for precise brain tumor segmentation in MRI images: A robust solution for medical image analysis," in 2024 International Conference on Signal Processing and Advance Research in Computing, 2024.
[18] X. Li, X. Fang, G. Yang, S. Su, L. Zhu, and Z. Yu, "TransU2-Net: an effective medical image segmentation framework based on transformer and U2-Net," IEEE Journal of Translational Engineering in Health and Medicine, vol. 11, pp. 441–450, 2023.
[19] A. Kumar, O. Murthy, Shrish, P. Ghosal, A. Mukherjee and D. Nandi, "A dense U-Net architecture for multiple sclerosis lesion segmentation," in IEEE Region 10 Conference, 2019.
[20] S. Aslani, M. Dayan, L. Storreli, 
M. Filippi,V. Murino and M. Rocca, "Multi-branch convolutional neural network for multiple sclerosis lesion segmentation," NeuroImage, vol. 196, pp. 1–15, 2019.
[21] M. E. Salem, S. Valverde and M. Cabezas, "Multiple sclerosis lesion synthesis in MRI using an encoder-decoder U-Net," IEEE Access, vol. 7, pp. 25171–25184, 2019.
[22] A. Alijamaat, A. NikravanShalmani, and P. Bayat, "Multiple sclerosis lesion segmentation from brain MRI using U-Net based on wavelet pooling," International Journal of Computer Assisted Radiology and Surgery, 2021.
[23] P. Ashtari, B. Barile, S. Van Huffel, and D. Sappey-Marinier, "New multiple sclerosis lesion segmentation and detection using pre-activation U-Net," Frontiers in Neuroscience, vol. 16, 2022.
[24] Z. Zhang, Q. Liu, and Y. Wang, "Road extraction by deep residual U-Net," IEEE Geoscience and Remote Sensing Letters, vol. 15, no. 5, pp. 749–753, 2018.
[25] J. Colman, L. Zhang, W. Duan, and X. Ye, "DR-Unet104 for multimodal MRI brain tumor segmentation," Lecture Notes in Computer Science, pp. 410–419, 2021.
[26] B. Sarica and D. Z. Seker, "New MS lesion segmentation with deep residual attention gate U-Net utilizing 2D slices of 3D MR images," Frontiers in Neuroscience, vol. 16, 2022.
[27] B. Sarica, D. Z. Seker, and B. Bayram, "A dense residual U-net for multiple sclerosis lesions segmentation from multi-sequence 3D MR images," International Journal of Medical Informatics, vol. 170, p. 104965, 2023.
[28] M. Rajasekar, A. Mukil, and R. Lakshamanan, "Segmentation and evaluation of multiple sclerosis in flair modality MRI with ResUNet," AIP conference proceedings, vol. 3161, 2024.
[29] V. Ganepola, P.  Mathur ,O. Adegboro, J. Dietlmeier and A. Lawlor "XAI-ResUNet: analysing the impact of pre-training in ResUNet architectures for multiple sclerosis lesion segmentation using EigenGradCAM," IET Conference Proceedings, no. 10, pp. 299–302, 2024
[30] O. Oktay, J. Schlemper, L. Folgoc, and M. Lee, "Attention U-Net: learning where to look for the pancreas," in 1st Conference on Medical Imaging with Deep Learning,,2018.
[31] C. hu, G. Kang, B. Hou, Y. Ma, F. Labeau and Z. Su, "ACU-NET: a 3D attention context U-Net for multiple sclerosis lesion segmentation," in IEEE International Conference on Acoustics, Speech and Signal Processing, 2020.
[32] M. Hashemi, M. Akhbari, and C. Jutten, "Delve into multiple sclerosis (MS) lesion exploration: A modified attention U-Net for MS lesion segmentation in brain MRI," Computers in Biology and Medicine, vol. 145, 2022.
[33] A. Rondinella , E. Crispino and F. Guarnera, "Boosting multiple sclerosis lesion segmentation through attention mechanism," Computers in Biology and Medicine, vol. 161, pp. 107021–107021, 2023.
[34] P. Ghosal, A. Roy and R. Agarwal, "Compound attention embedded dual channel encoder-decoder for ms lesion segmentation from brain MRI," Multimedia Tools and Applications, pp.1-33, 2024.
Applied and basic Machine intelligence research
Volume 3, Issue 1
August 2025
Pages 69-84

Files

Share

How to cite

Statistics