Multi-Scale Dilated Convolution Neural Network for Image Artifact Correction of Limited-Angle Tomography

被引:9
作者
Zhou, Haichuan [1 ,2 ]
Zhu, Yining [1 ,2 ]
Wang, Qian [3 ]
Xu, Jinqiu [4 ]
Li, Ge [5 ]
Chen, Defeng [1 ,2 ]
Dong, Yingying [6 ]
Zhang, Huitao [1 ,2 ]
机构
[1] Capital Normal Univ, Sch Math Sci, Beijing 100048, Peoples R China
[2] Capital Normal Univ, Beijing Adv Innovat Ctr Imaging Technol, Beijing 100048, Peoples R China
[3] Univ Massachusetts, Dept Elect & Comp Engn, Lowell, MA 01854 USA
[4] Paul Scherrer Inst, CH-5232 Villigen, Switzerland
[5] Ping An Technol, Beijing 100000, Peoples R China
[6] Chinese Acad Sci, Inst Remote Sensing & Digital Earth, Key Lab Digital Earth Sci, Beijing 100094, Peoples R China
基金
中国国家自然科学基金;
关键词
Limited-angle tomography; artifact correction; multi-scale; dilated convolution; COMPUTED-TOMOGRAPHY; RECONSTRUCTION; ART;
D O I
10.1109/ACCESS.2019.2962071
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Limited-angle computed tomography (CT) has arisen in some medical and industrial applications. It is also a challenging problem since some scan views are missing and the directly reconstructed images often suffer from severe distortions. For such kind of problems, we analyze the features of limited-angle CT images and propose a multi-scale dilated convolution neural network (MSD-CNN) to correct the artifacts and to restore the image. In this network, the dilated convolution layer and multi-scale pooling layer are combined to form a group and exited in the whole encoder-decoder process. Since the dilated convolutions support an exponential expansion of the receptive field without losing resolution and coverage, the obtained artifact features possess the multi-scale characteristic. Furthermore, to improve the effectiveness and accuracy of the training step, we employ a preprocessing method, which extracts image patches. Numerical experiments verify the out-performance of the proposed method compared with some conventional methods, such as Unet based deep learning,TV- and -based optimization methods.
引用
收藏
页码:1567 / 1576
页数:10
相关论文
共 35 条
[1]   Solving ill-posed inverse problems using iterative deep neural networks [J].
Adler, Jonas ;
Oktem, Ozan .
INVERSE PROBLEMS, 2017, 33 (12)
[2]   SIMULTANEOUS ALGEBRAIC RECONSTRUCTION TECHNIQUE (SART) - A SUPERIOR IMPLEMENTATION OF THE ART ALGORITHM [J].
ANDERSEN, AH ;
KAK, AC .
ULTRASONIC IMAGING, 1984, 6 (01) :81-94
[3]  
Anirudh R., 2018, ARXIV PREPRINT ARXIV
[4]  
Anirudh Rushil, 2017, ARXIV171110388
[5]  
[Anonymous], 1990, WAVELETS, DOI DOI 10.1007/978-3-642-75988-8_28
[6]  
[Anonymous], 2017, ARXIV170301382
[7]  
[Anonymous], 2016, arXiv Preprint, arXiv: 1603. 03805
[8]  
[Anonymous], 2016, P 9 ISCA WORKSH SPEE
[9]  
[Anonymous], 2012, ACM T GRAPH TOG
[10]  
[Anonymous], 2016, ARXIV160506409