An Automatic Glioma Segmentation System Based on A Separable Attention U-Net (SAUNet)

CORC > 沈阳自动化研究所 > 中国科学院沈阳自动化研究所 > 机器人学研究室

	An Automatic Glioma Segmentation System Based on A Separable Attention U-Net (SAUNet)
	Zhang, Zhenyu 1; Gao, Shouwei 1; Huang Z(黄钲)2
	2020
会议日期	October 16-18, 2020
会议地点	Virtual, Online, China
关键词	Glioma segmentation Separable attention U-Net
页码	95-101
英文摘要	With the complicated structure of brains, glioma segmentation is a challenging task. To precisely segment gliomas, U-Net structure is adopted by most current methods. However, the computation complexity of U-Net based method is large. Therefore, a separable attention U-Net, which can reduce the computation complexity without decreasing the performance, is proposed in this paper. Firstly, data augmentation techniques are implemented to enlarge the database and thus avoid over-fitting; Moreover, the separable attention U-Net is constructed for glioma segmentation. The experimental results indicate that the dice similarity coefficient of the proposed separable attention U-Net can reach 0.879 with the parameter number of 4.29 M, which indicates that the proposed glioma segmentation method is of application significance. A single column document that allows authors to type theircontent into the pre-existing set of paragraph formatting styles applied to the sample placeholder text here. Throughout the document you will find further instructions on how to format your text.
产权排序	2
会议录	Proceedings of 2020 9th International Conference on Bioinformatics and Biomedical Science, ICBBS 2020
会议录出版者	ACM
会议录出版地	New York
语种	英语
ISBN号	978-1-4503-8865-8
内容类型	会议论文
源URL	[http://ir.sia.cn/handle/173321/28293]
专题	沈阳自动化研究所_机器人学研究室
通讯作者	Zhang, Zhenyu
作者单位	1.School of Mechatronic Engineering and Automation, Shanghai University, Shanghai, China 2.Shenyang Institute of Automation Chinese Academy of Sciences, Shenyang, China
推荐引用方式 GB/T 7714	Zhang, Zhenyu,Gao, Shouwei,Huang Z. An Automatic Glioma Segmentation System Based on A Separable Attention U-Net (SAUNet)[C]. 见:. Virtual, Online, China. October 16-18, 2020.