U-Net-Based Models towards Optimal MR Brain Image Segmentation
Brain tumor segmentation from MRIs has always been a challenging task for radiologists, therefore, an automatic and generalized system to address this task is needed. Among all other deep learning techniques used in medical imaging, U-Net-based variants are the most used models found in the literatu...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-05-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/13/9/1624 |
| _version_ | 1797602758311804928 |
|---|---|
| author | Rammah Yousef Shakir Khan Gaurav Gupta Tamanna Siddiqui Bader M. Albahlal Saad Abdullah Alajlan Mohd Anul Haq |
| author_facet | Rammah Yousef Shakir Khan Gaurav Gupta Tamanna Siddiqui Bader M. Albahlal Saad Abdullah Alajlan Mohd Anul Haq |
| author_sort | Rammah Yousef |
| collection | DOAJ |
| description | Brain tumor segmentation from MRIs has always been a challenging task for radiologists, therefore, an automatic and generalized system to address this task is needed. Among all other deep learning techniques used in medical imaging, U-Net-based variants are the most used models found in the literature to segment medical images with respect to different modalities. Therefore, the goal of this paper is to examine the numerous advancements and innovations in the U-Net architecture, as well as recent trends, with the aim of highlighting the ongoing potential of U-Net being used to better the performance of brain tumor segmentation. Furthermore, we provide a quantitative comparison of different U-Net architectures to highlight the performance and the evolution of this network from an optimization perspective. In addition to that, we have experimented with four U-Net architectures (3D U-Net, Attention U-Net, R2 Attention U-Net, and modified 3D U-Net) on the BraTS 2020 dataset for brain tumor segmentation to provide a better overview of this architecture’s performance in terms of Dice score and Hausdorff distance 95%. Finally, we analyze the limitations and challenges of medical image analysis to provide a critical discussion about the importance of developing new architectures in terms of optimization. |
| first_indexed | 2024-03-11T04:21:09Z |
| format | Article |
| id | doaj.art-69dd237b7a1246b28a184700e73c6629 |
| institution | Directory Open Access Journal |
| issn | 2075-4418 |
| language | English |
| last_indexed | 2024-03-11T04:21:09Z |
| publishDate | 2023-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj.art-69dd237b7a1246b28a184700e73c66292023-11-17T22:46:13ZengMDPI AGDiagnostics2075-44182023-05-01139162410.3390/diagnostics13091624U-Net-Based Models towards Optimal MR Brain Image SegmentationRammah Yousef0Shakir Khan1Gaurav Gupta2Tamanna Siddiqui3Bader M. Albahlal4Saad Abdullah Alajlan5Mohd Anul Haq6Yogananda School of AI, Computers and Data Sciences, Shoolini University, Solan 173229, IndiaCollege of Computer and Information Sciences, Imam Mohammad Ibn Saud Islamic University (IMSIU), Riyadh 11432, Saudi ArabiaYogananda School of AI, Computers and Data Sciences, Shoolini University, Solan 173229, IndiaDepartment of Computer Science, Aligarh Muslim University, Aligarh 202001, IndiaCollege of Computer and Information Sciences, Imam Mohammad Ibn Saud Islamic University (IMSIU), Riyadh 11432, Saudi ArabiaCollege of Computer and Information Sciences, Imam Mohammad Ibn Saud Islamic University (IMSIU), Riyadh 11432, Saudi ArabiaDepartment of Computer Science, College of Computer and Information Sciences, Majmaah University, Al-Majmaah 11952, Saudi ArabiaBrain tumor segmentation from MRIs has always been a challenging task for radiologists, therefore, an automatic and generalized system to address this task is needed. Among all other deep learning techniques used in medical imaging, U-Net-based variants are the most used models found in the literature to segment medical images with respect to different modalities. Therefore, the goal of this paper is to examine the numerous advancements and innovations in the U-Net architecture, as well as recent trends, with the aim of highlighting the ongoing potential of U-Net being used to better the performance of brain tumor segmentation. Furthermore, we provide a quantitative comparison of different U-Net architectures to highlight the performance and the evolution of this network from an optimization perspective. In addition to that, we have experimented with four U-Net architectures (3D U-Net, Attention U-Net, R2 Attention U-Net, and modified 3D U-Net) on the BraTS 2020 dataset for brain tumor segmentation to provide a better overview of this architecture’s performance in terms of Dice score and Hausdorff distance 95%. Finally, we analyze the limitations and challenges of medical image analysis to provide a critical discussion about the importance of developing new architectures in terms of optimization.https://www.mdpi.com/2075-4418/13/9/1624U-NetMR brain imagesloss functionsimage segmentationoptimizationdeep learning |
| spellingShingle | Rammah Yousef Shakir Khan Gaurav Gupta Tamanna Siddiqui Bader M. Albahlal Saad Abdullah Alajlan Mohd Anul Haq U-Net-Based Models towards Optimal MR Brain Image Segmentation Diagnostics U-Net MR brain images loss functions image segmentation optimization deep learning |
| title | U-Net-Based Models towards Optimal MR Brain Image Segmentation |
| title_full | U-Net-Based Models towards Optimal MR Brain Image Segmentation |
| title_fullStr | U-Net-Based Models towards Optimal MR Brain Image Segmentation |
| title_full_unstemmed | U-Net-Based Models towards Optimal MR Brain Image Segmentation |
| title_short | U-Net-Based Models towards Optimal MR Brain Image Segmentation |
| title_sort | u net based models towards optimal mr brain image segmentation |
| topic | U-Net MR brain images loss functions image segmentation optimization deep learning |
| url | https://www.mdpi.com/2075-4418/13/9/1624 |
| work_keys_str_mv | AT rammahyousef unetbasedmodelstowardsoptimalmrbrainimagesegmentation AT shakirkhan unetbasedmodelstowardsoptimalmrbrainimagesegmentation AT gauravgupta unetbasedmodelstowardsoptimalmrbrainimagesegmentation AT tamannasiddiqui unetbasedmodelstowardsoptimalmrbrainimagesegmentation AT badermalbahlal unetbasedmodelstowardsoptimalmrbrainimagesegmentation AT saadabdullahalajlan unetbasedmodelstowardsoptimalmrbrainimagesegmentation AT mohdanulhaq unetbasedmodelstowardsoptimalmrbrainimagesegmentation |