Uncertainty-aware multi-resolution whole-body mr to ct synthesis

Kerstin Kläser; Pedro Borges; Richard Shaw; Marta Ranzini; Marc Modat; David Atkinson; Kris Thielemans; Brian Hutton; Vicky Goh; Gary Cook; M. Jorge Cardoso; Sébastien Ourselin

doi:10.1007/978-3-030-59520-3_12

Uncertainty-aware multi-resolution whole-body mr to ct synthesis

Kerstin Kläser^*, Pedro Borges, Richard Shaw, Marta Ranzini, Marc Modat, David Atkinson, Kris Thielemans, Brian Hutton, Vicky Goh, Gary Cook, M. Jorge Cardoso, Sébastien Ourselin

^*Corresponding author for this work

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

3 Citations (Scopus)

Abstract

Synthesising computed tomography (CT) images from magnetic resonance images (MRI) plays an important role in the field of medical image analysis, both for quantification and diagnostic purposes. Especially for brain applications, convolutional neural networks (CNNs) have proven to be a valuable tool in this image translation task, achieving state-of-the-art results. Full body image synthesis, however, remains largely uncharted territory, bearing many challenges including a limited field of view and large image size, complex spatial context and anatomical differences between time-elapsing image acquisitions. We propose a novel multi-resolution cascade 3D network for end-to-end full-body MR to CT synthesis. We show that our method outperforms popular CNNs like U-Net in 2D and 3D. We further propose to include uncertainty in our network as a measure of safety and to account for intrinsic noise and misalignment in the data.

Original language	English
Title of host publication	Simulation and Synthesis in Medical Imaging - 5th International Workshop, SASHIMI 2020, Held in Conjunction with MICCAI 2020, Proceedings
Editors	Ninon Burgos, David Svoboda, Jelmer M. Wolterink, Can Zhao
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	110-119
Number of pages	10
ISBN (Print)	9783030595197
DOIs	https://doi.org/10.1007/978-3-030-59520-3_12
Publication status	Published - 1 Jan 2020
Event	5th International Workshop on Simulation and Synthesis in Medical Imaging, SASHIMI 2020, held in conjunction with the Medical Image Computing and Computer Assisted Intervention, MICCAI 2020 - Lima, Peru Duration: 4 Oct 2020 → 4 Oct 2020

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12417 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	5th International Workshop on Simulation and Synthesis in Medical Imaging, SASHIMI 2020, held in conjunction with the Medical Image Computing and Computer Assisted Intervention, MICCAI 2020
Country/Territory	Peru
City	Lima
Period	4/10/2020 → 4/10/2020

Keywords

MR to CT synthesis
Multi-resolution CNN
Uncertainty

Access to Document

10.1007/978-3-030-59520-3_12

Cite this

Kläser, K., Borges, P., Shaw, R., Ranzini, M., Modat, M., Atkinson, D., Thielemans, K., Hutton, B., Goh, V., Cook, G., Cardoso, M. J., & Ourselin, S. (2020). Uncertainty-aware multi-resolution whole-body mr to ct synthesis. In N. Burgos, D. Svoboda, J. M. Wolterink, & C. Zhao (Eds.), Simulation and Synthesis in Medical Imaging - 5th International Workshop, SASHIMI 2020, Held in Conjunction with MICCAI 2020, Proceedings (pp. 110-119). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12417 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-59520-3_12

Kläser, Kerstin ; Borges, Pedro ; Shaw, Richard et al. / Uncertainty-aware multi-resolution whole-body mr to ct synthesis. Simulation and Synthesis in Medical Imaging - 5th International Workshop, SASHIMI 2020, Held in Conjunction with MICCAI 2020, Proceedings. editor / Ninon Burgos ; David Svoboda ; Jelmer M. Wolterink ; Can Zhao. Springer Science and Business Media Deutschland GmbH, 2020. pp. 110-119 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inbook{a26d358a6d6a4151b17d4f66296cca45,

title = "Uncertainty-aware multi-resolution whole-body mr to ct synthesis",

abstract = "Synthesising computed tomography (CT) images from magnetic resonance images (MRI) plays an important role in the field of medical image analysis, both for quantification and diagnostic purposes. Especially for brain applications, convolutional neural networks (CNNs) have proven to be a valuable tool in this image translation task, achieving state-of-the-art results. Full body image synthesis, however, remains largely uncharted territory, bearing many challenges including a limited field of view and large image size, complex spatial context and anatomical differences between time-elapsing image acquisitions. We propose a novel multi-resolution cascade 3D network for end-to-end full-body MR to CT synthesis. We show that our method outperforms popular CNNs like U-Net in 2D and 3D. We further propose to include uncertainty in our network as a measure of safety and to account for intrinsic noise and misalignment in the data.",

keywords = "MR to CT synthesis, Multi-resolution CNN, Uncertainty",

author = "Kerstin Kl{\"a}ser and Pedro Borges and Richard Shaw and Marta Ranzini and Marc Modat and David Atkinson and Kris Thielemans and Brian Hutton and Vicky Goh and Gary Cook and Cardoso, {M. Jorge} and S{\'e}bastien Ourselin",

year = "2020",

month = jan,

day = "1",

doi = "10.1007/978-3-030-59520-3_12",

language = "English",

isbn = "9783030595197",

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 = "110--119",

editor = "Ninon Burgos and David Svoboda and Wolterink, {Jelmer M.} and Can Zhao",

booktitle = "Simulation and Synthesis in Medical Imaging - 5th International Workshop, SASHIMI 2020, Held in Conjunction with MICCAI 2020, Proceedings",

address = "Germany",

note = "5th International Workshop on Simulation and Synthesis in Medical Imaging, SASHIMI 2020, held in conjunction with the Medical Image Computing and Computer Assisted Intervention, MICCAI 2020 ; Conference date: 04-10-2020 Through 04-10-2020",

}

Kläser, K, Borges, P, Shaw, R, Ranzini, M, Modat, M, Atkinson, D, Thielemans, K, Hutton, B, Goh, V , Cook, G , Cardoso, MJ & Ourselin, S 2020, Uncertainty-aware multi-resolution whole-body mr to ct synthesis. in N Burgos, D Svoboda, JM Wolterink & C Zhao (eds), Simulation and Synthesis in Medical Imaging - 5th International Workshop, SASHIMI 2020, Held in Conjunction with MICCAI 2020, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12417 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 110-119, 5th International Workshop on Simulation and Synthesis in Medical Imaging, SASHIMI 2020, held in conjunction with the Medical Image Computing and Computer Assisted Intervention, MICCAI 2020, Lima, Peru, 4/10/2020. https://doi.org/10.1007/978-3-030-59520-3_12

Uncertainty-aware multi-resolution whole-body mr to ct synthesis. / Kläser, Kerstin; Borges, Pedro; Shaw, Richard et al.
Simulation and Synthesis in Medical Imaging - 5th International Workshop, SASHIMI 2020, Held in Conjunction with MICCAI 2020, Proceedings. ed. / Ninon Burgos; David Svoboda; Jelmer M. Wolterink; Can Zhao. Springer Science and Business Media Deutschland GmbH, 2020. p. 110-119 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12417 LNCS).

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

TY - CHAP

T1 - Uncertainty-aware multi-resolution whole-body mr to ct synthesis

AU - Kläser, Kerstin

AU - Borges, Pedro

AU - Shaw, Richard

AU - Ranzini, Marta

AU - Modat, Marc

AU - Atkinson, David

AU - Thielemans, Kris

AU - Hutton, Brian

AU - Goh, Vicky

AU - Cook, Gary

AU - Cardoso, M. Jorge

AU - Ourselin, Sébastien

PY - 2020/1/1

Y1 - 2020/1/1

N2 - Synthesising computed tomography (CT) images from magnetic resonance images (MRI) plays an important role in the field of medical image analysis, both for quantification and diagnostic purposes. Especially for brain applications, convolutional neural networks (CNNs) have proven to be a valuable tool in this image translation task, achieving state-of-the-art results. Full body image synthesis, however, remains largely uncharted territory, bearing many challenges including a limited field of view and large image size, complex spatial context and anatomical differences between time-elapsing image acquisitions. We propose a novel multi-resolution cascade 3D network for end-to-end full-body MR to CT synthesis. We show that our method outperforms popular CNNs like U-Net in 2D and 3D. We further propose to include uncertainty in our network as a measure of safety and to account for intrinsic noise and misalignment in the data.

AB - Synthesising computed tomography (CT) images from magnetic resonance images (MRI) plays an important role in the field of medical image analysis, both for quantification and diagnostic purposes. Especially for brain applications, convolutional neural networks (CNNs) have proven to be a valuable tool in this image translation task, achieving state-of-the-art results. Full body image synthesis, however, remains largely uncharted territory, bearing many challenges including a limited field of view and large image size, complex spatial context and anatomical differences between time-elapsing image acquisitions. We propose a novel multi-resolution cascade 3D network for end-to-end full-body MR to CT synthesis. We show that our method outperforms popular CNNs like U-Net in 2D and 3D. We further propose to include uncertainty in our network as a measure of safety and to account for intrinsic noise and misalignment in the data.

KW - MR to CT synthesis

KW - Multi-resolution CNN

KW - Uncertainty

UR - http://www.scopus.com/inward/record.url?scp=85092176591&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-59520-3_12

DO - 10.1007/978-3-030-59520-3_12

M3 - Conference paper

AN - SCOPUS:85092176591

SN - 9783030595197

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

SP - 110

EP - 119

BT - Simulation and Synthesis in Medical Imaging - 5th International Workshop, SASHIMI 2020, Held in Conjunction with MICCAI 2020, Proceedings

A2 - Burgos, Ninon

A2 - Svoboda, David

A2 - Wolterink, Jelmer M.

A2 - Zhao, Can

PB - Springer Science and Business Media Deutschland GmbH

T2 - 5th International Workshop on Simulation and Synthesis in Medical Imaging, SASHIMI 2020, held in conjunction with the Medical Image Computing and Computer Assisted Intervention, MICCAI 2020

Y2 - 4 October 2020 through 4 October 2020

ER -

Kläser K, Borges P, Shaw R, Ranzini M, Modat M, Atkinson D et al. Uncertainty-aware multi-resolution whole-body mr to ct synthesis. In Burgos N, Svoboda D, Wolterink JM, Zhao C, editors, Simulation and Synthesis in Medical Imaging - 5th International Workshop, SASHIMI 2020, Held in Conjunction with MICCAI 2020, Proceedings. Springer Science and Business Media Deutschland GmbH. 2020. p. 110-119. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-59520-3_12

Uncertainty-aware multi-resolution whole-body mr to ct synthesis

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this