Cargando…

Unsupervised MR harmonization by learning disentangled representations using information bottleneck theory

In magnetic resonance (MR) imaging, a lack of standardization in acquisition often causes pulse sequence-based contrast variations in MR images from site to site, which impedes consistent measurements in automatic analyses In this paper, we propose an unsupervised MR image harmonization approach, CA...

Descripción completa

Detalles Bibliográficos
Autores principales: Zuo, Lianrui, Dewey, Blake E., Liu, Yihao, He, Yufan, Newsome, Scott D., Mowry, Ellen M., Resnick, Susan M., Prince, Jerry L., Carass, Aaron
Formato: Online Artículo Texto
Lenguaje:English
Publicado: 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10473284/
https://www.ncbi.nlm.nih.gov/pubmed/34506916
http://dx.doi.org/10.1016/j.neuroimage.2021.118569
_version_ 1785100243080052736
author Zuo, Lianrui
Dewey, Blake E.
Liu, Yihao
He, Yufan
Newsome, Scott D.
Mowry, Ellen M.
Resnick, Susan M.
Prince, Jerry L.
Carass, Aaron
author_facet Zuo, Lianrui
Dewey, Blake E.
Liu, Yihao
He, Yufan
Newsome, Scott D.
Mowry, Ellen M.
Resnick, Susan M.
Prince, Jerry L.
Carass, Aaron
author_sort Zuo, Lianrui
collection PubMed
description In magnetic resonance (MR) imaging, a lack of standardization in acquisition often causes pulse sequence-based contrast variations in MR images from site to site, which impedes consistent measurements in automatic analyses In this paper, we propose an unsupervised MR image harmonization approach, CALAMITI (Contrast Anatomy Learning and Analysis for MR Intensity Translation and Integration), which aims to alleviate contrast variations in multi-site MR imaging. Designed using information bottleneck theory, CALAMITI learns a globally disentangled latent space containing both anatomical and contrast information, which permits harmonization. In contrast to supervised harmonization methods, our approach does not need a sample population to be imaged across sites Unlike traditional unsupervised harmonization approaches which often suffer from geometry shifts, CALAMITI better preserves anatomy by design. The proposed method is also able to adapt to a new testing site with a straightforward fine-tuning process. Experiments on MR images acquired from ten sites show that CALAMITI achieves superior performance compared with other harmonization approaches.
format Online
Article
Text
id pubmed-10473284
institution National Center for Biotechnology Information
language English
publishDate 2021
record_format MEDLINE/PubMed
spelling pubmed-104732842023-09-01 Unsupervised MR harmonization by learning disentangled representations using information bottleneck theory Zuo, Lianrui Dewey, Blake E. Liu, Yihao He, Yufan Newsome, Scott D. Mowry, Ellen M. Resnick, Susan M. Prince, Jerry L. Carass, Aaron Neuroimage Article In magnetic resonance (MR) imaging, a lack of standardization in acquisition often causes pulse sequence-based contrast variations in MR images from site to site, which impedes consistent measurements in automatic analyses In this paper, we propose an unsupervised MR image harmonization approach, CALAMITI (Contrast Anatomy Learning and Analysis for MR Intensity Translation and Integration), which aims to alleviate contrast variations in multi-site MR imaging. Designed using information bottleneck theory, CALAMITI learns a globally disentangled latent space containing both anatomical and contrast information, which permits harmonization. In contrast to supervised harmonization methods, our approach does not need a sample population to be imaged across sites Unlike traditional unsupervised harmonization approaches which often suffer from geometry shifts, CALAMITI better preserves anatomy by design. The proposed method is also able to adapt to a new testing site with a straightforward fine-tuning process. Experiments on MR images acquired from ten sites show that CALAMITI achieves superior performance compared with other harmonization approaches. 2021-11 2021-09-08 /pmc/articles/PMC10473284/ /pubmed/34506916 http://dx.doi.org/10.1016/j.neuroimage.2021.118569 Text en https://creativecommons.org/licenses/by-nc-nd/4.0/This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/ (https://creativecommons.org/licenses/by-nc-nd/4.0/) )
spellingShingle Article
Zuo, Lianrui
Dewey, Blake E.
Liu, Yihao
He, Yufan
Newsome, Scott D.
Mowry, Ellen M.
Resnick, Susan M.
Prince, Jerry L.
Carass, Aaron
Unsupervised MR harmonization by learning disentangled representations using information bottleneck theory
title Unsupervised MR harmonization by learning disentangled representations using information bottleneck theory
title_full Unsupervised MR harmonization by learning disentangled representations using information bottleneck theory
title_fullStr Unsupervised MR harmonization by learning disentangled representations using information bottleneck theory
title_full_unstemmed Unsupervised MR harmonization by learning disentangled representations using information bottleneck theory
title_short Unsupervised MR harmonization by learning disentangled representations using information bottleneck theory
title_sort unsupervised mr harmonization by learning disentangled representations using information bottleneck theory
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10473284/
https://www.ncbi.nlm.nih.gov/pubmed/34506916
http://dx.doi.org/10.1016/j.neuroimage.2021.118569
work_keys_str_mv AT zuolianrui unsupervisedmrharmonizationbylearningdisentangledrepresentationsusinginformationbottlenecktheory
AT deweyblakee unsupervisedmrharmonizationbylearningdisentangledrepresentationsusinginformationbottlenecktheory
AT liuyihao unsupervisedmrharmonizationbylearningdisentangledrepresentationsusinginformationbottlenecktheory
AT heyufan unsupervisedmrharmonizationbylearningdisentangledrepresentationsusinginformationbottlenecktheory
AT newsomescottd unsupervisedmrharmonizationbylearningdisentangledrepresentationsusinginformationbottlenecktheory
AT mowryellenm unsupervisedmrharmonizationbylearningdisentangledrepresentationsusinginformationbottlenecktheory
AT resnicksusanm unsupervisedmrharmonizationbylearningdisentangledrepresentationsusinginformationbottlenecktheory
AT princejerryl unsupervisedmrharmonizationbylearningdisentangledrepresentationsusinginformationbottlenecktheory
AT carassaaron unsupervisedmrharmonizationbylearningdisentangledrepresentationsusinginformationbottlenecktheory