Joint Multi-anatomy Training of a Variational Network for Reconstruction of Accelerated Magnetic Resonance Image Acquisitions

Patricia M Johnson, Matthew J Muckley, Mary Bruno, Erich Kobler, Kerstin Hammernik, Thomas Pock, Florian Knoll

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Abstract

Magnetic resonance imaging is a leading image modality for many clinical applications; however, a significant drawback is the lengthy data acquisition. This motivates the development of methods for reconstruction of sparsely sampled image data. One such technique is the Variational Network (VN), a machine learning method that generalizes traditional iterative reconstruction techniques, learning the regularization term from large amounts of image data. Previously, with the VN technique, reconstruction of 4-fold accelerated knee images was shown to be highly successful. In this work we extend the VN approach to applications beyond knee imaging and evaluate the classic VN and a newly developed Unet-VN in 5 different anatomical regions. We evaluate the networks trained individually for each anatomical area as well as jointly trained with data from all anatomical areas. The VN and Unet-VN were …
Original languageEnglish
Title of host publicationMachine Learning for Medical Image Reconstruction
Subtitle of host publicationSecond International Workshop, MLMIR 2019, Held in Conjunction with MICCAI 2019, Shenzhen, China, October 17, 2019, Proceedings
EditorsF. Knoll, A. Maier, D. Rueckert, J. Ye
Place of PublicationCham
PublisherSpringer
Pages71-79
ISBN (Print)978-3-030-33842-8
DOIs
Publication statusPublished - 2019
Event2019 International Workshop on Machine Learning for Medical Image Reconstruction - Shenzen, China
Duration: 17 Oct 2019 → …

Publication series

NameLecture Notes in Computer Science
Volume11905

Conference

Conference2019 International Workshop on Machine Learning for Medical Image Reconstruction
Abbreviated titleMLMIR 2019
CountryChina
CityShenzen
Period17/10/19 → …

Fingerprint

Knee
Anatomy
Magnetic Resonance Spectroscopy
Joints
Magnetic Resonance Imaging
Learning
Machine Learning

Cite this

Johnson, P. M., Muckley, M. J., Bruno, M., Kobler, E., Hammernik, K., Pock, T., & Knoll, F. (2019). Joint Multi-anatomy Training of a Variational Network for Reconstruction of Accelerated Magnetic Resonance Image Acquisitions. In F. Knoll, A. Maier, D. Rueckert, & J. Ye (Eds.), Machine Learning for Medical Image Reconstruction: Second International Workshop, MLMIR 2019, Held in Conjunction with MICCAI 2019, Shenzhen, China, October 17, 2019, Proceedings (pp. 71-79). (Lecture Notes in Computer Science; Vol. 11905). Cham: Springer. https://doi.org/10.1007/978-3-030-33843-5_7

Joint Multi-anatomy Training of a Variational Network for Reconstruction of Accelerated Magnetic Resonance Image Acquisitions. / Johnson, Patricia M; Muckley, Matthew J; Bruno, Mary; Kobler, Erich; Hammernik, Kerstin; Pock, Thomas; Knoll, Florian.

Machine Learning for Medical Image Reconstruction: Second International Workshop, MLMIR 2019, Held in Conjunction with MICCAI 2019, Shenzhen, China, October 17, 2019, Proceedings. ed. / F. Knoll; A. Maier; D. Rueckert; J. Ye. Cham : Springer, 2019. p. 71-79 (Lecture Notes in Computer Science; Vol. 11905).

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Johnson, PM, Muckley, MJ, Bruno, M, Kobler, E, Hammernik, K, Pock, T & Knoll, F 2019, Joint Multi-anatomy Training of a Variational Network for Reconstruction of Accelerated Magnetic Resonance Image Acquisitions. in F Knoll, A Maier, D Rueckert & J Ye (eds), Machine Learning for Medical Image Reconstruction: Second International Workshop, MLMIR 2019, Held in Conjunction with MICCAI 2019, Shenzhen, China, October 17, 2019, Proceedings. Lecture Notes in Computer Science, vol. 11905, Springer, Cham, pp. 71-79, 2019 International Workshop on Machine Learning for Medical Image Reconstruction, Shenzen, China, 17/10/19. https://doi.org/10.1007/978-3-030-33843-5_7
Johnson PM, Muckley MJ, Bruno M, Kobler E, Hammernik K, Pock T et al. Joint Multi-anatomy Training of a Variational Network for Reconstruction of Accelerated Magnetic Resonance Image Acquisitions. In Knoll F, Maier A, Rueckert D, Ye J, editors, Machine Learning for Medical Image Reconstruction: Second International Workshop, MLMIR 2019, Held in Conjunction with MICCAI 2019, Shenzhen, China, October 17, 2019, Proceedings. Cham: Springer. 2019. p. 71-79. (Lecture Notes in Computer Science). https://doi.org/10.1007/978-3-030-33843-5_7
Johnson, Patricia M ; Muckley, Matthew J ; Bruno, Mary ; Kobler, Erich ; Hammernik, Kerstin ; Pock, Thomas ; Knoll, Florian. / Joint Multi-anatomy Training of a Variational Network for Reconstruction of Accelerated Magnetic Resonance Image Acquisitions. Machine Learning for Medical Image Reconstruction: Second International Workshop, MLMIR 2019, Held in Conjunction with MICCAI 2019, Shenzhen, China, October 17, 2019, Proceedings. editor / F. Knoll ; A. Maier ; D. Rueckert ; J. Ye. Cham : Springer, 2019. pp. 71-79 (Lecture Notes in Computer Science).
@inproceedings{0a9aac068fa1476fa60528079cd96c65,
title = "Joint Multi-anatomy Training of a Variational Network for Reconstruction of Accelerated Magnetic Resonance Image Acquisitions",
abstract = "Magnetic resonance imaging is a leading image modality for many clinical applications; however, a significant drawback is the lengthy data acquisition. This motivates the development of methods for reconstruction of sparsely sampled image data. One such technique is the Variational Network (VN), a machine learning method that generalizes traditional iterative reconstruction techniques, learning the regularization term from large amounts of image data. Previously, with the VN technique, reconstruction of 4-fold accelerated knee images was shown to be highly successful. In this work we extend the VN approach to applications beyond knee imaging and evaluate the classic VN and a newly developed Unet-VN in 5 different anatomical regions. We evaluate the networks trained individually for each anatomical area as well as jointly trained with data from all anatomical areas. The VN and Unet-VN were …",
author = "Johnson, {Patricia M} and Muckley, {Matthew J} and Mary Bruno and Erich Kobler and Kerstin Hammernik and Thomas Pock and Florian Knoll",
year = "2019",
doi = "10.1007/978-3-030-33843-5_7",
language = "English",
isbn = "978-3-030-33842-8",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "71--79",
editor = "F. Knoll and A. Maier and D. Rueckert and J. Ye",
booktitle = "Machine Learning for Medical Image Reconstruction",

}

TY - GEN

T1 - Joint Multi-anatomy Training of a Variational Network for Reconstruction of Accelerated Magnetic Resonance Image Acquisitions

AU - Johnson, Patricia M

AU - Muckley, Matthew J

AU - Bruno, Mary

AU - Kobler, Erich

AU - Hammernik, Kerstin

AU - Pock, Thomas

AU - Knoll, Florian

PY - 2019

Y1 - 2019

N2 - Magnetic resonance imaging is a leading image modality for many clinical applications; however, a significant drawback is the lengthy data acquisition. This motivates the development of methods for reconstruction of sparsely sampled image data. One such technique is the Variational Network (VN), a machine learning method that generalizes traditional iterative reconstruction techniques, learning the regularization term from large amounts of image data. Previously, with the VN technique, reconstruction of 4-fold accelerated knee images was shown to be highly successful. In this work we extend the VN approach to applications beyond knee imaging and evaluate the classic VN and a newly developed Unet-VN in 5 different anatomical regions. We evaluate the networks trained individually for each anatomical area as well as jointly trained with data from all anatomical areas. The VN and Unet-VN were …

AB - Magnetic resonance imaging is a leading image modality for many clinical applications; however, a significant drawback is the lengthy data acquisition. This motivates the development of methods for reconstruction of sparsely sampled image data. One such technique is the Variational Network (VN), a machine learning method that generalizes traditional iterative reconstruction techniques, learning the regularization term from large amounts of image data. Previously, with the VN technique, reconstruction of 4-fold accelerated knee images was shown to be highly successful. In this work we extend the VN approach to applications beyond knee imaging and evaluate the classic VN and a newly developed Unet-VN in 5 different anatomical regions. We evaluate the networks trained individually for each anatomical area as well as jointly trained with data from all anatomical areas. The VN and Unet-VN were …

U2 - 10.1007/978-3-030-33843-5_7

DO - 10.1007/978-3-030-33843-5_7

M3 - Conference contribution

SN - 978-3-030-33842-8

T3 - Lecture Notes in Computer Science

SP - 71

EP - 79

BT - Machine Learning for Medical Image Reconstruction

A2 - Knoll, F.

A2 - Maier, A.

A2 - Rueckert, D.

A2 - Ye, J.

PB - Springer

CY - Cham

ER -