Spatio-temporal regularization for deep MR Fingerprinting

Mohammad Golbabaee; Dongdong Chen; Mike E. Davies; Marion I. Menzel; Pedro A. Gómez

Spatio-temporal regularization for deep MR Fingerprinting

Mohammad Golbabaee, Dongdong Chen, Mike E. Davies, Marion I. Menzel, Pedro A. Gómez

School of Engineering

Research output: Contribution to conference › Paper › peer-review

Abstract

We study a deep learning approach to address the heavy storage and computation re- quirements of the baseline dictionary-matching (DM) for Magnetic Resonance Fingerprint- ing (MRF) reconstruction. The MRF-Net provides a piece-wise affine approximation to the (temporal) Bloch response manifold projection. Fed with non-iterated back-projected images, the network alone is unable to fully resolve spatially-correlated artefacts which ap- pear in highly undersampling regimes. We propose an accelerated iterative reconstruction to minimize these artefacts before feeding into the network. This is done through a convex regularization that jointly promotes spatio-temporal regularities of the MRF time-series.

Original language	Undefined/Unknown
Pages	1-4
Number of pages	4
Publication status	Published - 10 Jul 2019
Event	The 2nd International Conference on Medical Imaging with Deep Learning (MIDL) - Duration: 1 Jul 2019 → …

Conference

Conference	The 2nd International Conference on Medical Imaging with Deep Learning (MIDL)
Period	1/07/19 → …

Keywords / Materials (for Non-textual outputs)

magnetic resonance fingerprinting
deep learning
Regularisation

Access to Document

1809.01749v2Accepted author manuscript, 1.09 MB

https://openreview.net/forum?id=ryx64UL6YELicence: Creative Commons: Attribution (CC-BY)

C-SENSE: Exploiting low dimensional models in sensing, computation and signal processing
Davies, M. (Principal Investigator)
EU government bodies
1/09/16 → 31/08/22
Project: Research
CQ-MRI: Compressed Quantitative MRI
Marshall, I. (Principal Investigator) & Davies, M. (Co-investigator)
EPSRC
1/07/15 → 31/12/18
Project: Research

Cite this

@conference{b195c8445d364e56834e848b68ab820c,

title = "Spatio-temporal regularization for deep MR Fingerprinting",

abstract = "We study a deep learning approach to address the heavy storage and computation re- quirements of the baseline dictionary-matching (DM) for Magnetic Resonance Fingerprint- ing (MRF) reconstruction. The MRF-Net provides a piece-wise affine approximation to the (temporal) Bloch response manifold projection. Fed with non-iterated back-projected images, the network alone is unable to fully resolve spatially-correlated artefacts which ap- pear in highly undersampling regimes. We propose an accelerated iterative reconstruction to minimize these artefacts before feeding into the network. This is done through a convex regularization that jointly promotes spatio-temporal regularities of the MRF time-series.",

keywords = "magnetic resonance fingerprinting, deep learning, Regularisation",

author = "Mohammad Golbabaee and Dongdong Chen and Davies, {Mike E.} and Menzel, {Marion I.} and G{\'o}mez, {Pedro A.}",

year = "2019",

month = jul,

day = "10",

language = "Undefined/Unknown",

pages = "1--4",

note = "The 2nd International Conference on Medical Imaging with Deep Learning (MIDL) ; Conference date: 01-07-2019",

}

TY - CONF

T1 - Spatio-temporal regularization for deep MR Fingerprinting

AU - Golbabaee, Mohammad

AU - Chen, Dongdong

AU - Davies, Mike E.

AU - Menzel, Marion I.

AU - Gómez, Pedro A.

PY - 2019/7/10

Y1 - 2019/7/10

N2 - We study a deep learning approach to address the heavy storage and computation re- quirements of the baseline dictionary-matching (DM) for Magnetic Resonance Fingerprint- ing (MRF) reconstruction. The MRF-Net provides a piece-wise affine approximation to the (temporal) Bloch response manifold projection. Fed with non-iterated back-projected images, the network alone is unable to fully resolve spatially-correlated artefacts which ap- pear in highly undersampling regimes. We propose an accelerated iterative reconstruction to minimize these artefacts before feeding into the network. This is done through a convex regularization that jointly promotes spatio-temporal regularities of the MRF time-series.

AB - We study a deep learning approach to address the heavy storage and computation re- quirements of the baseline dictionary-matching (DM) for Magnetic Resonance Fingerprint- ing (MRF) reconstruction. The MRF-Net provides a piece-wise affine approximation to the (temporal) Bloch response manifold projection. Fed with non-iterated back-projected images, the network alone is unable to fully resolve spatially-correlated artefacts which ap- pear in highly undersampling regimes. We propose an accelerated iterative reconstruction to minimize these artefacts before feeding into the network. This is done through a convex regularization that jointly promotes spatio-temporal regularities of the MRF time-series.

KW - magnetic resonance fingerprinting

KW - deep learning

KW - Regularisation

M3 - Paper

SP - 1

EP - 4

T2 - The 2nd International Conference on Medical Imaging with Deep Learning (MIDL)

Y2 - 1 July 2019

ER -

Spatio-temporal regularization for deep MR Fingerprinting

Abstract

Conference

Keywords / Materials (for Non-textual outputs)

Access to Document

Projects

C-SENSE: Exploiting low dimensional models in sensing, computation and signal processing

CQ-MRI: Compressed Quantitative MRI

Cite this