InDuDoNet: An Interpretable Dual Domain Network for CT Metal Artifact Reduction

Hong Wang; Yuexiang Li; Haimiao Zhang; Jiawei Chen; Kai Ma; Deyu Meng; Yefeng Zheng

doi:10.1007/978-3-030-87231-1_11

InDuDoNet: An Interpretable Dual Domain Network for CT Metal Artifact Reduction

Hong Wang
, Yuexiang Li
, Haimiao Zhang
, Jiawei Chen
, Kai Ma
, Deyu Meng
, Yefeng Zheng

School of Mathematics and Statistics

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

65 Scopus citations

Abstract

For the task of metal artifact reduction (MAR), although deep learning (DL)-based methods have achieved promising performances, most of them suffer from two problems: 1) the CT imaging geometry constraint is not fully embedded into the network during training, leaving room for further performance improvement; 2) the model interpretability is lack of sufficient consideration. Against these issues, we propose a novel interpretable dual domain network, termed as InDuDoNet, which combines the advantages of model-driven and data-driven methodologies. Specifically, we build a joint spatial and Radon domain reconstruction model and utilize the proximal gradient technique to design an iterative algorithm for solving it. The optimization algorithm only consists of simple computational operators, which facilitate us to correspondingly unfold iterative steps into network modules and thus improve the interpretablility of the framework. Extensive experiments on synthesized and clinical data show the superiority of our InDuDoNet. Code is available in https://github.com/hongwang01/InDuDoNet.

Original language	English
Title of host publication	Medical Image Computing and Computer Assisted Intervention – MICCAI 2021 - 24th International Conference, Proceedings
Editors	Marleen de Bruijne, Marleen de Bruijne, Philippe C. Cattin, Stéphane Cotin, Nicolas Padoy, Stefanie Speidel, Yefeng Zheng, Caroline Essert
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	107-118
Number of pages	12
ISBN (Print)	9783030872304
DOIs	https://doi.org/10.1007/978-3-030-87231-1_11
State	Published - 2021
Event	24th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2021 - Virtual, Online Duration: 27 Sep 2021 → 1 Oct 2021

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12906 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	24th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2021
City	Virtual, Online
Period	27/09/21 → 1/10/21

Keywords

Generalization ability
Imaging geometry
Metal artifact reduction
Multi-class segmentation
Physical interpretability

Access to Document

10.1007/978-3-030-87231-1_11

Cite this

Wang, H., Li, Y., Zhang, H., Chen, J., Ma, K., Meng, D., & Zheng, Y. (2021). InDuDoNet: An Interpretable Dual Domain Network for CT Metal Artifact Reduction. In M. de Bruijne, M. de Bruijne, P. C. Cattin, S. Cotin, N. Padoy, S. Speidel, Y. Zheng, & C. Essert (Eds.), Medical Image Computing and Computer Assisted Intervention – MICCAI 2021 - 24th International Conference, Proceedings (pp. 107-118). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12906 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-87231-1_11

Wang, Hong ; Li, Yuexiang ; Zhang, Haimiao et al. / InDuDoNet : An Interpretable Dual Domain Network for CT Metal Artifact Reduction. Medical Image Computing and Computer Assisted Intervention – MICCAI 2021 - 24th International Conference, Proceedings. editor / Marleen de Bruijne ; Marleen de Bruijne ; Philippe C. Cattin ; Stéphane Cotin ; Nicolas Padoy ; Stefanie Speidel ; Yefeng Zheng ; Caroline Essert. Springer Science and Business Media Deutschland GmbH, 2021. pp. 107-118 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{0c817ce9d6854d008c002754016d9d7d,

title = "InDuDoNet: An Interpretable Dual Domain Network for CT Metal Artifact Reduction",

abstract = "For the task of metal artifact reduction (MAR), although deep learning (DL)-based methods have achieved promising performances, most of them suffer from two problems: 1) the CT imaging geometry constraint is not fully embedded into the network during training, leaving room for further performance improvement; 2) the model interpretability is lack of sufficient consideration. Against these issues, we propose a novel interpretable dual domain network, termed as InDuDoNet, which combines the advantages of model-driven and data-driven methodologies. Specifically, we build a joint spatial and Radon domain reconstruction model and utilize the proximal gradient technique to design an iterative algorithm for solving it. The optimization algorithm only consists of simple computational operators, which facilitate us to correspondingly unfold iterative steps into network modules and thus improve the interpretablility of the framework. Extensive experiments on synthesized and clinical data show the superiority of our InDuDoNet. Code is available in https://github.com/hongwang01/InDuDoNet.",

keywords = "Generalization ability, Imaging geometry, Metal artifact reduction, Multi-class segmentation, Physical interpretability",

author = "Hong Wang and Yuexiang Li and Haimiao Zhang and Jiawei Chen and Kai Ma and Deyu Meng and Yefeng Zheng",

note = "Publisher Copyright: {\textcopyright} 2021, Springer Nature Switzerland AG.; 24th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2021 ; Conference date: 27-09-2021 Through 01-10-2021",

year = "2021",

doi = "10.1007/978-3-030-87231-1\_11",

language = "英语",

isbn = "9783030872304",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "107--118",

editor = "\{de Bruijne\}, Marleen and \{de Bruijne\}, Marleen and Cattin, \{Philippe C.\} and St{\'e}phane Cotin and Nicolas Padoy and Stefanie Speidel and Yefeng Zheng and Caroline Essert",

booktitle = "Medical Image Computing and Computer Assisted Intervention – MICCAI 2021 - 24th International Conference, Proceedings",

}

Wang, H, Li, Y, Zhang, H, Chen, J, Ma, K, Meng, D & Zheng, Y 2021, InDuDoNet: An Interpretable Dual Domain Network for CT Metal Artifact Reduction. in M de Bruijne, M de Bruijne, PC Cattin, S Cotin, N Padoy, S Speidel, Y Zheng & C Essert (eds), Medical Image Computing and Computer Assisted Intervention – MICCAI 2021 - 24th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12906 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 107-118, 24th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2021, Virtual, Online, 27/09/21. https://doi.org/10.1007/978-3-030-87231-1_11

InDuDoNet: An Interpretable Dual Domain Network for CT Metal Artifact Reduction. / Wang, Hong; Li, Yuexiang; Zhang, Haimiao et al.
Medical Image Computing and Computer Assisted Intervention – MICCAI 2021 - 24th International Conference, Proceedings. ed. / Marleen de Bruijne; Marleen de Bruijne; Philippe C. Cattin; Stéphane Cotin; Nicolas Padoy; Stefanie Speidel; Yefeng Zheng; Caroline Essert. Springer Science and Business Media Deutschland GmbH, 2021. p. 107-118 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12906 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - InDuDoNet

T2 - 24th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2021

AU - Wang, Hong

AU - Li, Yuexiang

AU - Zhang, Haimiao

AU - Chen, Jiawei

AU - Ma, Kai

AU - Meng, Deyu

AU - Zheng, Yefeng

PY - 2021

Y1 - 2021

N2 - For the task of metal artifact reduction (MAR), although deep learning (DL)-based methods have achieved promising performances, most of them suffer from two problems: 1) the CT imaging geometry constraint is not fully embedded into the network during training, leaving room for further performance improvement; 2) the model interpretability is lack of sufficient consideration. Against these issues, we propose a novel interpretable dual domain network, termed as InDuDoNet, which combines the advantages of model-driven and data-driven methodologies. Specifically, we build a joint spatial and Radon domain reconstruction model and utilize the proximal gradient technique to design an iterative algorithm for solving it. The optimization algorithm only consists of simple computational operators, which facilitate us to correspondingly unfold iterative steps into network modules and thus improve the interpretablility of the framework. Extensive experiments on synthesized and clinical data show the superiority of our InDuDoNet. Code is available in https://github.com/hongwang01/InDuDoNet.

AB - For the task of metal artifact reduction (MAR), although deep learning (DL)-based methods have achieved promising performances, most of them suffer from two problems: 1) the CT imaging geometry constraint is not fully embedded into the network during training, leaving room for further performance improvement; 2) the model interpretability is lack of sufficient consideration. Against these issues, we propose a novel interpretable dual domain network, termed as InDuDoNet, which combines the advantages of model-driven and data-driven methodologies. Specifically, we build a joint spatial and Radon domain reconstruction model and utilize the proximal gradient technique to design an iterative algorithm for solving it. The optimization algorithm only consists of simple computational operators, which facilitate us to correspondingly unfold iterative steps into network modules and thus improve the interpretablility of the framework. Extensive experiments on synthesized and clinical data show the superiority of our InDuDoNet. Code is available in https://github.com/hongwang01/InDuDoNet.

KW - Generalization ability

KW - Imaging geometry

KW - Metal artifact reduction

KW - Multi-class segmentation

KW - Physical interpretability

UR - https://www.scopus.com/pages/publications/85116489169

U2 - 10.1007/978-3-030-87231-1_11

DO - 10.1007/978-3-030-87231-1_11

M3 - 会议稿件

AN - SCOPUS:85116489169

SN - 9783030872304

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 107

EP - 118

BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2021 - 24th International Conference, Proceedings

A2 - de Bruijne, Marleen

A2 - Cattin, Philippe C.

A2 - Cotin, Stéphane

A2 - Padoy, Nicolas

A2 - Speidel, Stefanie

A2 - Zheng, Yefeng

A2 - Essert, Caroline

PB - Springer Science and Business Media Deutschland GmbH

Y2 - 27 September 2021 through 1 October 2021

ER -

Wang H, Li Y, Zhang H, Chen J, Ma K, Meng D et al. InDuDoNet: An Interpretable Dual Domain Network for CT Metal Artifact Reduction. In de Bruijne M, de Bruijne M, Cattin PC, Cotin S, Padoy N, Speidel S, Zheng Y, Essert C, editors, Medical Image Computing and Computer Assisted Intervention – MICCAI 2021 - 24th International Conference, Proceedings. Springer Science and Business Media Deutschland GmbH. 2021. p. 107-118. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-87231-1_11

InDuDoNet: An Interpretable Dual Domain Network for CT Metal Artifact Reduction

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this