Semi-Supervised Cerebrovascular Segmentation by Hierarchical Convolutional Neural Network
Due to the tortuosity and the complexity of cerebral vasculature and the similar intensity distribution with the background, it remains challenging to accurately segment cerebral vessels from magnetic resonance angiography (MRA). The previous rule-based methods have limitations when applied to accur...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2018-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/8522028/ |
| _version_ | 1818927722228875264 |
|---|---|
| author | Fengjun Zhao Yibing Chen Fei Chen Xuelei He Xin Cao Yuqing Hou Huangjian Yi Xiaowei He Jimin Liang |
| author_facet | Fengjun Zhao Yibing Chen Fei Chen Xuelei He Xin Cao Yuqing Hou Huangjian Yi Xiaowei He Jimin Liang |
| author_sort | Fengjun Zhao |
| collection | DOAJ |
| description | Due to the tortuosity and the complexity of cerebral vasculature and the similar intensity distribution with the background, it remains challenging to accurately segment cerebral vessels from magnetic resonance angiography (MRA). The previous rule-based methods have limitations when applied to accurate clinical diagnosis, such as the under-segmentation on complex vessels, the dependence on domain knowledge, and the lack of quantification estimation. In this paper, we proposed a semi-supervised cerebrovascular segmentation method with a hierarchical convolutional neural network (H-CNN) that transfers the exquisite model/feature design in rule-based methods to solve the mapping from MRA images to cerebral vessels. First, we generated the tube-level labels of cerebral vessels with centerlines and estimated radii. Second, we constructed and trained an H-CNN with the MRA images and corresponding tube-level labels. Third, the stopping criterion of the proposed H-CNN was determined by the comprehensive index (CI) that was defined based on partially annotated voxel-level ground truth. The comparison of our H-CNN with the vesselness, bi-Gaussian, optimally oriented flux, vessel enhancing diffusion, hybrid diffusion with continuous switch, Mimics software, convolutional neutral network<sub>2D</sub> (CNN)<sub>2D</sub>, and CNN<sub>3D</sub> were conducted on six testing images. The mean sensitivity, accuracy, and the CI of our H-CNN are 94.69%, 97.85%, and 2.99%, respectively, outperforming the other methods. The curved planar reformation also visualized the performance of H-CNN for cerebrovascular segmentation. Given only the tube-level labels, the proposed H-CNN method accomplished the voxel-level vessel segmentation via the hierarchical update of CNN. The H-CNN can potentially to be applied for the accurate diagnosis of cerebrovascular diseases and other medical image segmentation with only partially correct labels. |
| first_indexed | 2024-12-20T03:17:32Z |
| format | Article |
| id | doaj.art-511ae50f69e743909b9a3f047b4ef494 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-20T03:17:32Z |
| publishDate | 2018-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-511ae50f69e743909b9a3f047b4ef4942022-12-21T19:55:19ZengIEEEIEEE Access2169-35362018-01-016678416785210.1109/ACCESS.2018.28795218522028Semi-Supervised Cerebrovascular Segmentation by Hierarchical Convolutional Neural NetworkFengjun Zhao0https://orcid.org/0000-0001-8658-8412Yibing Chen1Fei Chen2Xuelei He3Xin Cao4Yuqing Hou5Huangjian Yi6Xiaowei He7https://orcid.org/0000-0003-2126-178XJimin Liang8School of Information Sciences and Technology, Northwest University, Xi’an, ChinaSchool of Information Sciences and Technology, Northwest University, Xi’an, ChinaSchool of Life Science and Technology, Xidian University, Xi’an, ChinaSchool of Information Sciences and Technology, Northwest University, Xi’an, ChinaSchool of Information Sciences and Technology, Northwest University, Xi’an, ChinaSchool of Information Sciences and Technology, Northwest University, Xi’an, ChinaSchool of Information Sciences and Technology, Northwest University, Xi’an, ChinaSchool of Information Sciences and Technology, Northwest University, Xi’an, ChinaSchool of Life Science and Technology, Xidian University, Xi’an, ChinaDue to the tortuosity and the complexity of cerebral vasculature and the similar intensity distribution with the background, it remains challenging to accurately segment cerebral vessels from magnetic resonance angiography (MRA). The previous rule-based methods have limitations when applied to accurate clinical diagnosis, such as the under-segmentation on complex vessels, the dependence on domain knowledge, and the lack of quantification estimation. In this paper, we proposed a semi-supervised cerebrovascular segmentation method with a hierarchical convolutional neural network (H-CNN) that transfers the exquisite model/feature design in rule-based methods to solve the mapping from MRA images to cerebral vessels. First, we generated the tube-level labels of cerebral vessels with centerlines and estimated radii. Second, we constructed and trained an H-CNN with the MRA images and corresponding tube-level labels. Third, the stopping criterion of the proposed H-CNN was determined by the comprehensive index (CI) that was defined based on partially annotated voxel-level ground truth. The comparison of our H-CNN with the vesselness, bi-Gaussian, optimally oriented flux, vessel enhancing diffusion, hybrid diffusion with continuous switch, Mimics software, convolutional neutral network<sub>2D</sub> (CNN)<sub>2D</sub>, and CNN<sub>3D</sub> were conducted on six testing images. The mean sensitivity, accuracy, and the CI of our H-CNN are 94.69%, 97.85%, and 2.99%, respectively, outperforming the other methods. The curved planar reformation also visualized the performance of H-CNN for cerebrovascular segmentation. Given only the tube-level labels, the proposed H-CNN method accomplished the voxel-level vessel segmentation via the hierarchical update of CNN. The H-CNN can potentially to be applied for the accurate diagnosis of cerebrovascular diseases and other medical image segmentation with only partially correct labels.https://ieeexplore.ieee.org/document/8522028/Magnetic resonance angiographycerebral blood vesselcenterlinesegmentationconvolutional neural network |
| spellingShingle | Fengjun Zhao Yibing Chen Fei Chen Xuelei He Xin Cao Yuqing Hou Huangjian Yi Xiaowei He Jimin Liang Semi-Supervised Cerebrovascular Segmentation by Hierarchical Convolutional Neural Network IEEE Access Magnetic resonance angiography cerebral blood vessel centerline segmentation convolutional neural network |
| title | Semi-Supervised Cerebrovascular Segmentation by Hierarchical Convolutional Neural Network |
| title_full | Semi-Supervised Cerebrovascular Segmentation by Hierarchical Convolutional Neural Network |
| title_fullStr | Semi-Supervised Cerebrovascular Segmentation by Hierarchical Convolutional Neural Network |
| title_full_unstemmed | Semi-Supervised Cerebrovascular Segmentation by Hierarchical Convolutional Neural Network |
| title_short | Semi-Supervised Cerebrovascular Segmentation by Hierarchical Convolutional Neural Network |
| title_sort | semi supervised cerebrovascular segmentation by hierarchical convolutional neural network |
| topic | Magnetic resonance angiography cerebral blood vessel centerline segmentation convolutional neural network |
| url | https://ieeexplore.ieee.org/document/8522028/ |
| work_keys_str_mv | AT fengjunzhao semisupervisedcerebrovascularsegmentationbyhierarchicalconvolutionalneuralnetwork AT yibingchen semisupervisedcerebrovascularsegmentationbyhierarchicalconvolutionalneuralnetwork AT feichen semisupervisedcerebrovascularsegmentationbyhierarchicalconvolutionalneuralnetwork AT xueleihe semisupervisedcerebrovascularsegmentationbyhierarchicalconvolutionalneuralnetwork AT xincao semisupervisedcerebrovascularsegmentationbyhierarchicalconvolutionalneuralnetwork AT yuqinghou semisupervisedcerebrovascularsegmentationbyhierarchicalconvolutionalneuralnetwork AT huangjianyi semisupervisedcerebrovascularsegmentationbyhierarchicalconvolutionalneuralnetwork AT xiaoweihe semisupervisedcerebrovascularsegmentationbyhierarchicalconvolutionalneuralnetwork AT jiminliang semisupervisedcerebrovascularsegmentationbyhierarchicalconvolutionalneuralnetwork |