RA-UNet: A Hybrid Deep Attention-Aware Network to Extract Liver and Tumor in CT Scans
Automatic extraction of liver and tumor from CT volumes is a challenging task due to their heterogeneous and diffusive shapes. Recently, 2D deep convolutional neural networks have become popular in medical image segmentation tasks because of the utilization of large labeled datasets to learn hierarc...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2020-12-01
|
| Series: | Frontiers in Bioengineering and Biotechnology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fbioe.2020.605132/full |
| _version_ | 1818595982944763904 |
|---|---|
| author | Qiangguo Jin Qiangguo Jin Zhaopeng Meng Zhaopeng Meng Changming Sun Hui Cui Ran Su |
| author_facet | Qiangguo Jin Qiangguo Jin Zhaopeng Meng Zhaopeng Meng Changming Sun Hui Cui Ran Su |
| author_sort | Qiangguo Jin |
| collection | DOAJ |
| description | Automatic extraction of liver and tumor from CT volumes is a challenging task due to their heterogeneous and diffusive shapes. Recently, 2D deep convolutional neural networks have become popular in medical image segmentation tasks because of the utilization of large labeled datasets to learn hierarchical features. However, few studies investigate 3D networks for liver tumor segmentation. In this paper, we propose a 3D hybrid residual attention-aware segmentation method, i.e., RA-UNet, to precisely extract the liver region and segment tumors from the liver. The proposed network has a basic architecture as U-Net which extracts contextual information combining low-level feature maps with high-level ones. Attention residual modules are integrated so that the attention-aware features change adaptively. This is the first work that an attention residual mechanism is used to segment tumors from 3D medical volumetric images. We evaluated our framework on the public MICCAI 2017 Liver Tumor Segmentation dataset and tested the generalization on the 3DIRCADb dataset. The experiments show that our architecture obtains competitive results. |
| first_indexed | 2024-12-16T11:24:40Z |
| format | Article |
| id | doaj.art-4d387d15936740dbbe066027e63f9c0f |
| institution | Directory Open Access Journal |
| issn | 2296-4185 |
| language | English |
| last_indexed | 2024-12-16T11:24:40Z |
| publishDate | 2020-12-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Bioengineering and Biotechnology |
| spelling | doaj.art-4d387d15936740dbbe066027e63f9c0f2022-12-21T22:33:23ZengFrontiers Media S.A.Frontiers in Bioengineering and Biotechnology2296-41852020-12-01810.3389/fbioe.2020.605132605132RA-UNet: A Hybrid Deep Attention-Aware Network to Extract Liver and Tumor in CT ScansQiangguo Jin0Qiangguo Jin1Zhaopeng Meng2Zhaopeng Meng3Changming Sun4Hui Cui5Ran Su6School of Computer Software, College of Intelligence and Computing, Tianjin University, Tianjin, ChinaCSIRO Data61, Sydney, NSW, AustraliaSchool of Computer Software, College of Intelligence and Computing, Tianjin University, Tianjin, ChinaTianjin University of Traditional Chinese Medicine, Tianjin, ChinaCSIRO Data61, Sydney, NSW, AustraliaDepartment of Computer Science and Information Technology, La Trobe University, Melbourne, VIC, AustraliaSchool of Computer Software, College of Intelligence and Computing, Tianjin University, Tianjin, ChinaAutomatic extraction of liver and tumor from CT volumes is a challenging task due to their heterogeneous and diffusive shapes. Recently, 2D deep convolutional neural networks have become popular in medical image segmentation tasks because of the utilization of large labeled datasets to learn hierarchical features. However, few studies investigate 3D networks for liver tumor segmentation. In this paper, we propose a 3D hybrid residual attention-aware segmentation method, i.e., RA-UNet, to precisely extract the liver region and segment tumors from the liver. The proposed network has a basic architecture as U-Net which extracts contextual information combining low-level feature maps with high-level ones. Attention residual modules are integrated so that the attention-aware features change adaptively. This is the first work that an attention residual mechanism is used to segment tumors from 3D medical volumetric images. We evaluated our framework on the public MICCAI 2017 Liver Tumor Segmentation dataset and tested the generalization on the 3DIRCADb dataset. The experiments show that our architecture obtains competitive results.https://www.frontiersin.org/articles/10.3389/fbioe.2020.605132/fullmedical image segmentationtumor segmentationu-netresidual learningattention mechanism |
| spellingShingle | Qiangguo Jin Qiangguo Jin Zhaopeng Meng Zhaopeng Meng Changming Sun Hui Cui Ran Su RA-UNet: A Hybrid Deep Attention-Aware Network to Extract Liver and Tumor in CT Scans Frontiers in Bioengineering and Biotechnology medical image segmentation tumor segmentation u-net residual learning attention mechanism |
| title | RA-UNet: A Hybrid Deep Attention-Aware Network to Extract Liver and Tumor in CT Scans |
| title_full | RA-UNet: A Hybrid Deep Attention-Aware Network to Extract Liver and Tumor in CT Scans |
| title_fullStr | RA-UNet: A Hybrid Deep Attention-Aware Network to Extract Liver and Tumor in CT Scans |
| title_full_unstemmed | RA-UNet: A Hybrid Deep Attention-Aware Network to Extract Liver and Tumor in CT Scans |
| title_short | RA-UNet: A Hybrid Deep Attention-Aware Network to Extract Liver and Tumor in CT Scans |
| title_sort | ra unet a hybrid deep attention aware network to extract liver and tumor in ct scans |
| topic | medical image segmentation tumor segmentation u-net residual learning attention mechanism |
| url | https://www.frontiersin.org/articles/10.3389/fbioe.2020.605132/full |
| work_keys_str_mv | AT qiangguojin raunetahybriddeepattentionawarenetworktoextractliverandtumorinctscans AT qiangguojin raunetahybriddeepattentionawarenetworktoextractliverandtumorinctscans AT zhaopengmeng raunetahybriddeepattentionawarenetworktoextractliverandtumorinctscans AT zhaopengmeng raunetahybriddeepattentionawarenetworktoextractliverandtumorinctscans AT changmingsun raunetahybriddeepattentionawarenetworktoextractliverandtumorinctscans AT huicui raunetahybriddeepattentionawarenetworktoextractliverandtumorinctscans AT ransu raunetahybriddeepattentionawarenetworktoextractliverandtumorinctscans |