| |||||||
ShanghaiTech University Knowledge Management System
MetricUNet: Synergistic image- and voxel-level learning for precise prostate segmentation via online sampling | |
2021-07 | |
发表期刊 | MEDICAL IMAGE ANALYSIS (IF:10.7[JCR-2023],11.9[5-Year]) |
ISSN | 1361-8415 |
EISSN | 1361-8423 |
卷号 | 71 |
发表状态 | 已发表 |
DOI | 10.1016/j.media.2021.102039 |
摘要 | Fully convolutional networks (FCNs), including UNet and VNet, are widely-used network architectures for semantic segmentation in recent studies. However, conventional FCN is typically trained by the cross entropy or Dice loss, which only calculates the error between predictions and ground-truth labels for pixels individually. This often results in non-smooth neighborhoods in the predicted segmentation. This problem becomes more serious in CT prostate segmentation as CT images are usually of low tissue contrast. To address this problem, we propose a two-stage framework, with the first stage to quickly localize the prostate region, and the second stage to precisely segment the prostate by a multi-task UNet architecture. We introduce a novel online metric learning module through voxel-wise sampling in the multitask network. Therefore, the proposed network has a dual-branch architecture that tackles two tasks: (1) a segmentation sub-network aiming to generate the prostate segmentation, and (2) a voxel-metric learning sub-network aiming to improve the quality of the learned feature space supervised by a metric loss. Specifically, the voxel-metric learning sub-network samples tuples (including triplets and pairs) in voxel-level through the intermediate feature maps. Unlike conventional deep metric learning methods that generate triplets or pairs in image-level before the training phase, our proposed voxel-wise tuples are sampled in an online manner and operated in an end-to-end fashion via multi-task learning. To evaluate the proposed method, we implement extensive experiments on a real CT image dataset consisting 339 patients. The ablation studies show that our method can effectively learn more representative voxellevel features compared with the conventional learning methods with cross-entropy or Dice loss. And the comparisons show that the proposed method outperforms the state-of-the-art methods by a reasonable margin. (c) 2021 Elsevier B.V. All rights reserved. |
关键词 | Prostate cancer Sampling Metric learning Fully convolutional networks Triplet Contrast learning |
收录类别 | SCIE ; EI |
语种 | 英语 |
WOS研究方向 | Computer Science ; Engineering ; Radiology, Nuclear Medicine & Medical Imaging |
WOS类目 | Computer Science, Artificial Intelligence ; Computer Science, Interdisciplinary Applications ; Engineering, Biomedical ; Radiology, Nuclear Medicine & Medical Imaging |
WOS记录号 | WOS:000663615600013 |
出版者 | ELSEVIER |
原始文献类型 | Article |
引用统计 | 正在获取...
|
文献类型 | 期刊论文 |
条目标识符 | https://kms.shanghaitech.edu.cn/handle/2MSLDSTB/127598 |
专题 | 生物医学工程学院_PI研究组_沈定刚组 |
通讯作者 | Zhang, Junfeng; Shen, Dinggang |
作者单位 | 1.Nanjing Univ, Med Sch, Nanjing, Peoples R China; 2.Nanjing Univ, Natl Inst Healthcare Data Sci, Nanjing, Peoples R China; 3.Xi An Jiao Tong Univ, Sch Math & Stat, Xian, Shanxi, Peoples R China; 4.Stanford Univ, Dept Psychiat & Behav Sci, Stanford, CA 94305 USA; 5.Stanford Univ, Dept Comp Sci, Stanford, CA 94305 USA; 6.Nanjing Univ, State Key Lab Novel Software Technol, Nanjing, Peoples R China; 7.Nanjing Univ, Sch Med, Nanjing Drum Tower Hosp, Dept Radiol, Nanjing, Peoples R China; 8.ShanghaiTech Univ, Sch Biomed Engn, Shanghai, Peoples R China; 9.Shanghai United Imaging Intelligence Co Ltd, Dept Res & Dev, Shanghai, Peoples R China; 10.Korea Univ, Dept Artificial Intelligence, Seoul 02841, South Korea |
通讯作者单位 | 生物医学工程学院 |
推荐引用方式 GB/T 7714 | He, Kelei,Lian, Chunfeng,Adeli, Ehsan,et al. MetricUNet: Synergistic image- and voxel-level learning for precise prostate segmentation via online sampling[J]. MEDICAL IMAGE ANALYSIS,2021,71. |
APA | He, Kelei.,Lian, Chunfeng.,Adeli, Ehsan.,Huo, Jing.,Gao, Yang.,...&Shen, Dinggang.(2021).MetricUNet: Synergistic image- and voxel-level learning for precise prostate segmentation via online sampling.MEDICAL IMAGE ANALYSIS,71. |
MLA | He, Kelei,et al."MetricUNet: Synergistic image- and voxel-level learning for precise prostate segmentation via online sampling".MEDICAL IMAGE ANALYSIS 71(2021). |
条目包含的文件 | ||||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 |
修改评论
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。