Low-field magnetic resonance image enhancement via stochastic image quality transfer

Hongxiang Lin; Matteo Figini; Felice D'Arco; Godwin Ogbole; Ryutaro Tanno; Stefano B. Blumberg; Lisa Ronan; Biobele J. Brown; David W. Carmichael; Ikeoluwa Lagunju; Judith Helen Cross; Delmiro Fernandez-Reyes; Daniel C. Alexander

doi:10.1016/j.media.2023.102807

Low-field magnetic resonance image enhancement via stochastic image quality transfer

Hongxiang Lin^*, Matteo Figini, Felice D'Arco, Godwin Ogbole, Ryutaro Tanno, Stefano B. Blumberg, Lisa Ronan, Biobele J. Brown, David W. Carmichael, Ikeoluwa Lagunju, Judith Helen Cross, Delmiro Fernandez-Reyes, Daniel C. Alexander

^*Corresponding author for this work

Biomedical Engineering Department

Research output: Contribution to journal › Article › peer-review

Abstract

Low-field (<1T) magnetic resonance imaging (MRI) scanners remain in widespread use in low- and middle-income countries (LMICs) and are commonly used for some applications in higher income countries e.g. for small child patients with obesity, claustrophobia, implants, or tattoos. However, low-field MR images commonly have lower resolution and poorer contrast than images from high field (1.5T, 3T, and above). Here, we present Image Quality Transfer (IQT) to enhance low-field structural MRI by estimating from a low-field image the image we would have obtained from the same subject at high field. Our approach uses (i) a stochastic low-field image simulator as the forward model to capture uncertainty and variation in the contrast of low-field images corresponding to a particular high-field image, and (ii) an anisotropic U-Net variant specifically designed for the IQT inverse problem. We evaluate the proposed algorithm both in simulation and using multi-contrast (T1-weighted, T2-weighted, and fluid attenuated inversion recovery (FLAIR)) clinical low-field MRI data from an LMIC hospital. We show the efficacy of IQT in improving contrast and resolution of low-field MR images. We demonstrate that IQT-enhanced images have potential for enhancing visualisation of anatomical structures and pathological lesions of clinical relevance from the perspective of radiologists. IQT is proved to have capability of boosting the diagnostic value of low-field MRI, especially in low-resource settings.

Original language	English
Article number	102807
Journal	Medical Image Analysis
Volume	87
DOIs	https://doi.org/10.1016/j.media.2023.102807
Publication status	Published - Jul 2023

Keywords

Deep neural networks
Image quality transfer
Low-field MRI
Stochastic simulator

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1016/j.media.2023.102807

Cite this

@article{aa380a1a6c45484f9c840d491a60a325,

title = "Low-field magnetic resonance image enhancement via stochastic image quality transfer",

abstract = "Low-field (<1T) magnetic resonance imaging (MRI) scanners remain in widespread use in low- and middle-income countries (LMICs) and are commonly used for some applications in higher income countries e.g. for small child patients with obesity, claustrophobia, implants, or tattoos. However, low-field MR images commonly have lower resolution and poorer contrast than images from high field (1.5T, 3T, and above). Here, we present Image Quality Transfer (IQT) to enhance low-field structural MRI by estimating from a low-field image the image we would have obtained from the same subject at high field. Our approach uses (i) a stochastic low-field image simulator as the forward model to capture uncertainty and variation in the contrast of low-field images corresponding to a particular high-field image, and (ii) an anisotropic U-Net variant specifically designed for the IQT inverse problem. We evaluate the proposed algorithm both in simulation and using multi-contrast (T1-weighted, T2-weighted, and fluid attenuated inversion recovery (FLAIR)) clinical low-field MRI data from an LMIC hospital. We show the efficacy of IQT in improving contrast and resolution of low-field MR images. We demonstrate that IQT-enhanced images have potential for enhancing visualisation of anatomical structures and pathological lesions of clinical relevance from the perspective of radiologists. IQT is proved to have capability of boosting the diagnostic value of low-field MRI, especially in low-resource settings.",

keywords = "Deep neural networks, Image quality transfer, Low-field MRI, Stochastic simulator",

author = "Hongxiang Lin and Matteo Figini and Felice D'Arco and Godwin Ogbole and Ryutaro Tanno and Blumberg, {Stefano B.} and Lisa Ronan and Brown, {Biobele J.} and Carmichael, {David W.} and Ikeoluwa Lagunju and Cross, {Judith Helen} and Delmiro Fernandez-Reyes and Alexander, {Daniel C.}",

note = "Funding Information: This work was supported by EPSRC, United Kingdom grants ( EP/R014019/1 , EP/R006032/1 and EP/M020533/1 ), the NIHR UCLH Biomedical Research Centre, United Kingdom , and the NIHR Biomedical Research Centre at Great Ormond Street Hospital, United Kingdom . 3T T1w and T2w images were provided in part by the Human Connectome Project, WU-Minn Consortium (Principal Investigators: David Van Essen and Kamil Ugurbil; 1U54MH091657 ) funded by the 16 NIH Institutes and Centers that support the NIH Blueprint for Neuroscience Research, United States ; and by the McDonnell Center for Systems Neuroscience at Washington University, United States . High-field FLAIR images were used in part from the Leipzig Study for Mind-Body-Emotion Interactions (LEMON) data set provided by the Mind-Body-Emotion group at the Max Planck Institute for Human Cognitive and Brain Sciences. For the purpose of open access, the author has applied a Creative Commons Attribution (CC BY) licence to any Author Accepted Manuscript version arising. Publisher Copyright: {\textcopyright} 2023 Elsevier B.V.",

year = "2023",

month = jul,

doi = "10.1016/j.media.2023.102807",

language = "English",

volume = "87",

journal = "Medical Image Analysis",

issn = "1361-8415",

publisher = "Elsevier",

}

TY - JOUR

T1 - Low-field magnetic resonance image enhancement via stochastic image quality transfer

AU - Lin, Hongxiang

AU - Figini, Matteo

AU - D'Arco, Felice

AU - Ogbole, Godwin

AU - Tanno, Ryutaro

AU - Blumberg, Stefano B.

AU - Ronan, Lisa

AU - Brown, Biobele J.

AU - Carmichael, David W.

AU - Lagunju, Ikeoluwa

AU - Cross, Judith Helen

AU - Fernandez-Reyes, Delmiro

AU - Alexander, Daniel C.

N1 - Funding Information: This work was supported by EPSRC, United Kingdom grants ( EP/R014019/1 , EP/R006032/1 and EP/M020533/1 ), the NIHR UCLH Biomedical Research Centre, United Kingdom , and the NIHR Biomedical Research Centre at Great Ormond Street Hospital, United Kingdom . 3T T1w and T2w images were provided in part by the Human Connectome Project, WU-Minn Consortium (Principal Investigators: David Van Essen and Kamil Ugurbil; 1U54MH091657 ) funded by the 16 NIH Institutes and Centers that support the NIH Blueprint for Neuroscience Research, United States ; and by the McDonnell Center for Systems Neuroscience at Washington University, United States . High-field FLAIR images were used in part from the Leipzig Study for Mind-Body-Emotion Interactions (LEMON) data set provided by the Mind-Body-Emotion group at the Max Planck Institute for Human Cognitive and Brain Sciences. For the purpose of open access, the author has applied a Creative Commons Attribution (CC BY) licence to any Author Accepted Manuscript version arising. Publisher Copyright: © 2023 Elsevier B.V.

PY - 2023/7

Y1 - 2023/7

N2 - Low-field (<1T) magnetic resonance imaging (MRI) scanners remain in widespread use in low- and middle-income countries (LMICs) and are commonly used for some applications in higher income countries e.g. for small child patients with obesity, claustrophobia, implants, or tattoos. However, low-field MR images commonly have lower resolution and poorer contrast than images from high field (1.5T, 3T, and above). Here, we present Image Quality Transfer (IQT) to enhance low-field structural MRI by estimating from a low-field image the image we would have obtained from the same subject at high field. Our approach uses (i) a stochastic low-field image simulator as the forward model to capture uncertainty and variation in the contrast of low-field images corresponding to a particular high-field image, and (ii) an anisotropic U-Net variant specifically designed for the IQT inverse problem. We evaluate the proposed algorithm both in simulation and using multi-contrast (T1-weighted, T2-weighted, and fluid attenuated inversion recovery (FLAIR)) clinical low-field MRI data from an LMIC hospital. We show the efficacy of IQT in improving contrast and resolution of low-field MR images. We demonstrate that IQT-enhanced images have potential for enhancing visualisation of anatomical structures and pathological lesions of clinical relevance from the perspective of radiologists. IQT is proved to have capability of boosting the diagnostic value of low-field MRI, especially in low-resource settings.

AB - Low-field (<1T) magnetic resonance imaging (MRI) scanners remain in widespread use in low- and middle-income countries (LMICs) and are commonly used for some applications in higher income countries e.g. for small child patients with obesity, claustrophobia, implants, or tattoos. However, low-field MR images commonly have lower resolution and poorer contrast than images from high field (1.5T, 3T, and above). Here, we present Image Quality Transfer (IQT) to enhance low-field structural MRI by estimating from a low-field image the image we would have obtained from the same subject at high field. Our approach uses (i) a stochastic low-field image simulator as the forward model to capture uncertainty and variation in the contrast of low-field images corresponding to a particular high-field image, and (ii) an anisotropic U-Net variant specifically designed for the IQT inverse problem. We evaluate the proposed algorithm both in simulation and using multi-contrast (T1-weighted, T2-weighted, and fluid attenuated inversion recovery (FLAIR)) clinical low-field MRI data from an LMIC hospital. We show the efficacy of IQT in improving contrast and resolution of low-field MR images. We demonstrate that IQT-enhanced images have potential for enhancing visualisation of anatomical structures and pathological lesions of clinical relevance from the perspective of radiologists. IQT is proved to have capability of boosting the diagnostic value of low-field MRI, especially in low-resource settings.

KW - Deep neural networks

KW - Image quality transfer

KW - Low-field MRI

KW - Stochastic simulator

UR - http://www.scopus.com/inward/record.url?scp=85153683220&partnerID=8YFLogxK

U2 - 10.1016/j.media.2023.102807

DO - 10.1016/j.media.2023.102807

M3 - Article

AN - SCOPUS:85153683220

SN - 1361-8415

VL - 87

JO - Medical Image Analysis

JF - Medical Image Analysis

M1 - 102807

ER -

Low-field magnetic resonance image enhancement via stochastic image quality transfer

Abstract

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this