Translation of 1D inverse fourier transform of k-space to an image based on deep learning for accelerating magnetic resonance imaging

Taejoon Eo, Hyungseob Shin, Taeseong Kim, Yohan Jun, Do Sik Hwang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

To reconstruct magnetic resonance (MR) images from undersampled Cartesian k-space data, we propose an algorithm based on two deep-learning architectures: (1) a multi-layer perceptron (MLP) that estimates a target image from 1D inverse Fourier transform (IFT) of k-space; and (2) a convolutional neural network (CNN) that estimates the target image from the estimated image of the MLP. The MLP learns the relationship between 1D IFT of undersampled k-space which is transformed along the frequency-encoding direction and the target fully-sampled image. The MLP is trained line by line rather than by a whole image, because each frequency-encoding line of the 1D IFT of k-space is not correlated with each other. It can dramatically decrease the number of parameters to be learned because the number of input/output pixels decrease from N2 to N. The next CNN learns the relationship between an estimated image of the MLP and the target fully-sampled image to reduce remaining artifacts in the image domain. The proposed deep-learning algorithm (i.e., the combination of the MLP and the CNN) exhibited superior performance over a single MLP and a single CNN. And it outperformed the comparison algorithms including CS-MRI, DL-MRI, a CNN-based algorithm (denoted as Wang’s algorithm), PANO, and FDLCP in both qualitative and quantitative evaluation. Consequently, the proposed algorithm is applicable up to a sampling ratio of 25% in Cartesian k-space.

Original languageEnglish
Title of host publicationMedical Image Computing and Computer Assisted Intervention – MICCAI 2018 - 21st International Conference, 2018, Proceedings
EditorsJulia A. Schnabel, Christos Davatzikos, Carlos Alberola-López, Gabor Fichtinger, Alejandro F. Frangi
PublisherSpringer Verlag
Pages241-249
Number of pages9
ISBN (Print)9783030009274
DOIs
Publication statusPublished - 2018 Jan 1
Event21st International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2018 - Granada, Spain
Duration: 2018 Sep 162018 Sep 20

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11070 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other21st International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2018
CountrySpain
CityGranada
Period18/9/1618/9/20

Fingerprint

Inverse transforms
K-space
Magnetic Resonance Imaging
Multilayer neural networks
Magnetic resonance
Perceptron
Fourier transform
Fourier transforms
Multilayer
Imaging techniques
Neural networks
Neural Networks
Target
Magnetic resonance imaging
Cartesian
Line
Encoding
Decrease
Magnetic Resonance Image
Learning

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Eo, T., Shin, H., Kim, T., Jun, Y., & Hwang, D. S. (2018). Translation of 1D inverse fourier transform of k-space to an image based on deep learning for accelerating magnetic resonance imaging. In J. A. Schnabel, C. Davatzikos, C. Alberola-López, G. Fichtinger, & A. F. Frangi (Eds.), Medical Image Computing and Computer Assisted Intervention – MICCAI 2018 - 21st International Conference, 2018, Proceedings (pp. 241-249). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11070 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-00928-1_28
Eo, Taejoon ; Shin, Hyungseob ; Kim, Taeseong ; Jun, Yohan ; Hwang, Do Sik. / Translation of 1D inverse fourier transform of k-space to an image based on deep learning for accelerating magnetic resonance imaging. Medical Image Computing and Computer Assisted Intervention – MICCAI 2018 - 21st International Conference, 2018, Proceedings. editor / Julia A. Schnabel ; Christos Davatzikos ; Carlos Alberola-López ; Gabor Fichtinger ; Alejandro F. Frangi. Springer Verlag, 2018. pp. 241-249 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{1700f07fbc27402db8274c4eb07e69dc,
title = "Translation of 1D inverse fourier transform of k-space to an image based on deep learning for accelerating magnetic resonance imaging",
abstract = "To reconstruct magnetic resonance (MR) images from undersampled Cartesian k-space data, we propose an algorithm based on two deep-learning architectures: (1) a multi-layer perceptron (MLP) that estimates a target image from 1D inverse Fourier transform (IFT) of k-space; and (2) a convolutional neural network (CNN) that estimates the target image from the estimated image of the MLP. The MLP learns the relationship between 1D IFT of undersampled k-space which is transformed along the frequency-encoding direction and the target fully-sampled image. The MLP is trained line by line rather than by a whole image, because each frequency-encoding line of the 1D IFT of k-space is not correlated with each other. It can dramatically decrease the number of parameters to be learned because the number of input/output pixels decrease from N2 to N. The next CNN learns the relationship between an estimated image of the MLP and the target fully-sampled image to reduce remaining artifacts in the image domain. The proposed deep-learning algorithm (i.e., the combination of the MLP and the CNN) exhibited superior performance over a single MLP and a single CNN. And it outperformed the comparison algorithms including CS-MRI, DL-MRI, a CNN-based algorithm (denoted as Wang’s algorithm), PANO, and FDLCP in both qualitative and quantitative evaluation. Consequently, the proposed algorithm is applicable up to a sampling ratio of 25{\%} in Cartesian k-space.",
author = "Taejoon Eo and Hyungseob Shin and Taeseong Kim and Yohan Jun and Hwang, {Do Sik}",
year = "2018",
month = "1",
day = "1",
doi = "10.1007/978-3-030-00928-1_28",
language = "English",
isbn = "9783030009274",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "241--249",
editor = "Schnabel, {Julia A.} and Christos Davatzikos and Carlos Alberola-L{\'o}pez and Gabor Fichtinger and Frangi, {Alejandro F.}",
booktitle = "Medical Image Computing and Computer Assisted Intervention – MICCAI 2018 - 21st International Conference, 2018, Proceedings",
address = "Germany",

}

Eo, T, Shin, H, Kim, T, Jun, Y & Hwang, DS 2018, Translation of 1D inverse fourier transform of k-space to an image based on deep learning for accelerating magnetic resonance imaging. in JA Schnabel, C Davatzikos, C Alberola-López, G Fichtinger & AF Frangi (eds), Medical Image Computing and Computer Assisted Intervention – MICCAI 2018 - 21st International Conference, 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11070 LNCS, Springer Verlag, pp. 241-249, 21st International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2018, Granada, Spain, 18/9/16. https://doi.org/10.1007/978-3-030-00928-1_28

Translation of 1D inverse fourier transform of k-space to an image based on deep learning for accelerating magnetic resonance imaging. / Eo, Taejoon; Shin, Hyungseob; Kim, Taeseong; Jun, Yohan; Hwang, Do Sik.

Medical Image Computing and Computer Assisted Intervention – MICCAI 2018 - 21st International Conference, 2018, Proceedings. ed. / Julia A. Schnabel; Christos Davatzikos; Carlos Alberola-López; Gabor Fichtinger; Alejandro F. Frangi. Springer Verlag, 2018. p. 241-249 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11070 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Translation of 1D inverse fourier transform of k-space to an image based on deep learning for accelerating magnetic resonance imaging

AU - Eo, Taejoon

AU - Shin, Hyungseob

AU - Kim, Taeseong

AU - Jun, Yohan

AU - Hwang, Do Sik

PY - 2018/1/1

Y1 - 2018/1/1

N2 - To reconstruct magnetic resonance (MR) images from undersampled Cartesian k-space data, we propose an algorithm based on two deep-learning architectures: (1) a multi-layer perceptron (MLP) that estimates a target image from 1D inverse Fourier transform (IFT) of k-space; and (2) a convolutional neural network (CNN) that estimates the target image from the estimated image of the MLP. The MLP learns the relationship between 1D IFT of undersampled k-space which is transformed along the frequency-encoding direction and the target fully-sampled image. The MLP is trained line by line rather than by a whole image, because each frequency-encoding line of the 1D IFT of k-space is not correlated with each other. It can dramatically decrease the number of parameters to be learned because the number of input/output pixels decrease from N2 to N. The next CNN learns the relationship between an estimated image of the MLP and the target fully-sampled image to reduce remaining artifacts in the image domain. The proposed deep-learning algorithm (i.e., the combination of the MLP and the CNN) exhibited superior performance over a single MLP and a single CNN. And it outperformed the comparison algorithms including CS-MRI, DL-MRI, a CNN-based algorithm (denoted as Wang’s algorithm), PANO, and FDLCP in both qualitative and quantitative evaluation. Consequently, the proposed algorithm is applicable up to a sampling ratio of 25% in Cartesian k-space.

AB - To reconstruct magnetic resonance (MR) images from undersampled Cartesian k-space data, we propose an algorithm based on two deep-learning architectures: (1) a multi-layer perceptron (MLP) that estimates a target image from 1D inverse Fourier transform (IFT) of k-space; and (2) a convolutional neural network (CNN) that estimates the target image from the estimated image of the MLP. The MLP learns the relationship between 1D IFT of undersampled k-space which is transformed along the frequency-encoding direction and the target fully-sampled image. The MLP is trained line by line rather than by a whole image, because each frequency-encoding line of the 1D IFT of k-space is not correlated with each other. It can dramatically decrease the number of parameters to be learned because the number of input/output pixels decrease from N2 to N. The next CNN learns the relationship between an estimated image of the MLP and the target fully-sampled image to reduce remaining artifacts in the image domain. The proposed deep-learning algorithm (i.e., the combination of the MLP and the CNN) exhibited superior performance over a single MLP and a single CNN. And it outperformed the comparison algorithms including CS-MRI, DL-MRI, a CNN-based algorithm (denoted as Wang’s algorithm), PANO, and FDLCP in both qualitative and quantitative evaluation. Consequently, the proposed algorithm is applicable up to a sampling ratio of 25% in Cartesian k-space.

UR - http://www.scopus.com/inward/record.url?scp=85054073010&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85054073010&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-00928-1_28

DO - 10.1007/978-3-030-00928-1_28

M3 - Conference contribution

AN - SCOPUS:85054073010

SN - 9783030009274

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

SP - 241

EP - 249

BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2018 - 21st International Conference, 2018, Proceedings

A2 - Schnabel, Julia A.

A2 - Davatzikos, Christos

A2 - Alberola-López, Carlos

A2 - Fichtinger, Gabor

A2 - Frangi, Alejandro F.

PB - Springer Verlag

ER -

Eo T, Shin H, Kim T, Jun Y, Hwang DS. Translation of 1D inverse fourier transform of k-space to an image based on deep learning for accelerating magnetic resonance imaging. In Schnabel JA, Davatzikos C, Alberola-López C, Fichtinger G, Frangi AF, editors, Medical Image Computing and Computer Assisted Intervention – MICCAI 2018 - 21st International Conference, 2018, Proceedings. Springer Verlag. 2018. p. 241-249. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-00928-1_28