Dual-Attention-Based Enhanced Unified Net for Precise GTV Segmentation of Nasopharyngeal Carcinoma in 3D MR Images
Accurate gross tumor volume (GTV) segmentation is essential for effective radiotherapy in nasopharyngeal carcinoma (NPC). However, challenges arise due to the nasopharyngeal region’s complex anatomy and the annotated data scarcity. Our study presents a dual-attention-based enhanced unifie...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11037670/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849420387220193280 |
|---|---|
| author | Hassan Ali Khan Gong Xueqing Muhammad Shoib Amin Zeeshan Bin Siddique Muhammad Ahtsam Naeem |
| author_facet | Hassan Ali Khan Gong Xueqing Muhammad Shoib Amin Zeeshan Bin Siddique Muhammad Ahtsam Naeem |
| author_sort | Hassan Ali Khan |
| collection | DOAJ |
| description | Accurate gross tumor volume (GTV) segmentation is essential for effective radiotherapy in nasopharyngeal carcinoma (NPC). However, challenges arise due to the nasopharyngeal region’s complex anatomy and the annotated data scarcity. Our study presents a dual-attention-based enhanced unified network (DAEU-Net) designed for precise NPC GTV segmentation utilizing 3D T1, T2, and T1C-weighted MR images. Our approach involves splitting large-scale MR data into multiple patches and then training every patch independently. This approach effectively captures localized and detailed information without downscaling the image resolution. The DAEU-Net integrates channel-attention and pixel-attention modules within the encoder section, eliminating background noise and reducing information loss by enhancing the network’s focus on detailed features. The decoder section incorporates bottleneck residual blocks to enhance the computing efficiency and robustness of the network. The proposed methodology surpasses the state-of-the-art models with a respective average symmetric surface distance (ASSD) of <inline-formula> <tex-math notation="LaTeX">$0.920\pm 0.386$ </tex-math></inline-formula> mm, <inline-formula> <tex-math notation="LaTeX">$0.987\pm 0.421$ </tex-math></inline-formula> mm, and <inline-formula> <tex-math notation="LaTeX">$1.043\pm 0.457$ </tex-math></inline-formula> mm and a dice similarity coefficient (DSC) of 0.896, 0.871, and 0.851, respectively. Multi-viewed animated MR images in three orthogonal dimensions (axial, sagittal, and coronal) with predicted NPC tumors and real GTV masks were shown to assist in comprehending the tumor’s precise location. Our approach can significantly improves NPC tumor delineation, aids in automated tumor lesion segmentation, and reduces the annotation workload for oncologists. |
| format | Article |
| id | doaj-art-a3671eff2c1d472aabf0a3f7d1f6819b |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-a3671eff2c1d472aabf0a3f7d1f6819b2025-08-20T03:31:46ZengIEEEIEEE Access2169-35362025-01-011312303112304010.1109/ACCESS.2025.358060011037670Dual-Attention-Based Enhanced Unified Net for Precise GTV Segmentation of Nasopharyngeal Carcinoma in 3D MR ImagesHassan Ali Khan0https://orcid.org/0000-0002-3311-363XGong Xueqing1https://orcid.org/0000-0001-7532-7153Muhammad Shoib Amin2https://orcid.org/0000-0001-9046-1502Zeeshan Bin Siddique3https://orcid.org/0000-0002-3208-4064Muhammad Ahtsam Naeem4Software Engineering Institute, East China Normal University, Shanghai, ChinaSoftware Engineering Institute, East China Normal University, Shanghai, ChinaSoftware Engineering Institute, East China Normal University, Shanghai, ChinaCollege of Computer Science and Technology, Qingdao University, Qingdao, ChinaSchool of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu, Sichuan, ChinaAccurate gross tumor volume (GTV) segmentation is essential for effective radiotherapy in nasopharyngeal carcinoma (NPC). However, challenges arise due to the nasopharyngeal region’s complex anatomy and the annotated data scarcity. Our study presents a dual-attention-based enhanced unified network (DAEU-Net) designed for precise NPC GTV segmentation utilizing 3D T1, T2, and T1C-weighted MR images. Our approach involves splitting large-scale MR data into multiple patches and then training every patch independently. This approach effectively captures localized and detailed information without downscaling the image resolution. The DAEU-Net integrates channel-attention and pixel-attention modules within the encoder section, eliminating background noise and reducing information loss by enhancing the network’s focus on detailed features. The decoder section incorporates bottleneck residual blocks to enhance the computing efficiency and robustness of the network. The proposed methodology surpasses the state-of-the-art models with a respective average symmetric surface distance (ASSD) of <inline-formula> <tex-math notation="LaTeX">$0.920\pm 0.386$ </tex-math></inline-formula> mm, <inline-formula> <tex-math notation="LaTeX">$0.987\pm 0.421$ </tex-math></inline-formula> mm, and <inline-formula> <tex-math notation="LaTeX">$1.043\pm 0.457$ </tex-math></inline-formula> mm and a dice similarity coefficient (DSC) of 0.896, 0.871, and 0.851, respectively. Multi-viewed animated MR images in three orthogonal dimensions (axial, sagittal, and coronal) with predicted NPC tumors and real GTV masks were shown to assist in comprehending the tumor’s precise location. Our approach can significantly improves NPC tumor delineation, aids in automated tumor lesion segmentation, and reduces the annotation workload for oncologists.https://ieeexplore.ieee.org/document/11037670/Segmentationmedical imagingNPCneural networksU-Netdeep learning |
| spellingShingle | Hassan Ali Khan Gong Xueqing Muhammad Shoib Amin Zeeshan Bin Siddique Muhammad Ahtsam Naeem Dual-Attention-Based Enhanced Unified Net for Precise GTV Segmentation of Nasopharyngeal Carcinoma in 3D MR Images IEEE Access Segmentation medical imaging NPC neural networks U-Net deep learning |
| title | Dual-Attention-Based Enhanced Unified Net for Precise GTV Segmentation of Nasopharyngeal Carcinoma in 3D MR Images |
| title_full | Dual-Attention-Based Enhanced Unified Net for Precise GTV Segmentation of Nasopharyngeal Carcinoma in 3D MR Images |
| title_fullStr | Dual-Attention-Based Enhanced Unified Net for Precise GTV Segmentation of Nasopharyngeal Carcinoma in 3D MR Images |
| title_full_unstemmed | Dual-Attention-Based Enhanced Unified Net for Precise GTV Segmentation of Nasopharyngeal Carcinoma in 3D MR Images |
| title_short | Dual-Attention-Based Enhanced Unified Net for Precise GTV Segmentation of Nasopharyngeal Carcinoma in 3D MR Images |
| title_sort | dual attention based enhanced unified net for precise gtv segmentation of nasopharyngeal carcinoma in 3d mr images |
| topic | Segmentation medical imaging NPC neural networks U-Net deep learning |
| url | https://ieeexplore.ieee.org/document/11037670/ |
| work_keys_str_mv | AT hassanalikhan dualattentionbasedenhancedunifiednetforprecisegtvsegmentationofnasopharyngealcarcinomain3dmrimages AT gongxueqing dualattentionbasedenhancedunifiednetforprecisegtvsegmentationofnasopharyngealcarcinomain3dmrimages AT muhammadshoibamin dualattentionbasedenhancedunifiednetforprecisegtvsegmentationofnasopharyngealcarcinomain3dmrimages AT zeeshanbinsiddique dualattentionbasedenhancedunifiednetforprecisegtvsegmentationofnasopharyngealcarcinomain3dmrimages AT muhammadahtsamnaeem dualattentionbasedenhancedunifiednetforprecisegtvsegmentationofnasopharyngealcarcinomain3dmrimages |