Deep Image Priors for Magnetic Resonance Fingerprinting with Pretrained Bloch-Consistent Denoising Autoencoders

Perla Jazmin Mayo Diaz de Leon; Matteo Cencini; Ketan Fatania; Carolin Pirkl; Marion Menzel; Bjoern Menze; Michela Tosetti; Mohammad Golbabaee

doi:10.1109/ISBI56570.2024.10635677

Deep Image Priors for Magnetic Resonance Fingerprinting with Pretrained Bloch-Consistent Denoising Autoencoders

Perla Jazmin Mayo Diaz de Leon, Matteo Cencini, Ketan Fatania, Carolin Pirkl, Marion Menzel, Bjoern Menze, Michela Tosetti, Mohammad Golbabaee

School of Engineering Mathematics and Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

8 Downloads (Pure)

Abstract

The estimation of multi-parametric quantitative maps from Magnetic Resonance Fingerprinting (MRF) compressed sampled acquisitions, albeit successful, remains a challenge due to the high underspampling rate and artifacts naturally occuring during image reconstruction. Whilst state-of-the-art DL methods can successfully address the task, to fully exploit their capabilities they often require training on a paired dataset, in an area where ground truth is seldom available. In this work, we propose a method that combines a deep image prior (DIP) module that, without ground truth and in conjunction with a Bloch consistency enforcing autoencoder, can tackle the problem, resulting in a method faster and of equivalent or better accuracy than DIP-MRF.

Original language	English
Title of host publication	2024 IEEE 21st International Symposium on Biomedical Imaging (ISBI)
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Number of pages	4
ISBN (Electronic)	9798350313338
ISBN (Print)	9798350313345
DOIs	https://doi.org/10.1109/ISBI56570.2024.10635677
Publication status	Published - 22 Aug 2024
Event	21st IEEE International Symposium on Biomedical Imaging - Megaron Athens International Conference Center, Athens, Greece Duration: 27 May 2024 → 30 May 2024 https://biomedicalimaging.org/2024/

Publication series

Name	Proceedings of the IEEE International Symposium on Biomedical Imaging
Publisher	IEEE
ISSN (Print)	1945-7928
ISSN (Electronic)	1945-8452

Conference

Conference	21st IEEE International Symposium on Biomedical Imaging
Abbreviated title	ISBI 2024
Country/Territory	Greece
City	Athens
Period	27/05/24 → 30/05/24
Internet address	https://biomedicalimaging.org/2024/

Bibliographical note

Publisher Copyright:
© 2024 IEEE.

Keywords

magnetic resonance fingerprinting
deep learning
deep image priors
quantitative magnetic resonance imaging

Access to Document

10.1109/ISBI56570.2024.10635677

Full-text PDF (accepted manuscript)
This is the accepted author manuscript (AAM) of the article which has been made Open Access under the University of Bristol's Scholarly Works Policy. The final published version (Version of Record) can be found on the publisher's website. The copyright of any third-party content, such as images, remains with the copyright holder.
Accepted author manuscript, 3.04 MBLicence: CC BY

https://arxiv.org/abs/2407.19866

Handle.net

Persistent link

Cite this

Mayo Diaz de Leon, P. J., Cencini, M., Fatania, K., Pirkl, C., Menzel, M., Menze, B., Tosetti, M., & Golbabaee, M. (2024). Deep Image Priors for Magnetic Resonance Fingerprinting with Pretrained Bloch-Consistent Denoising Autoencoders. In 2024 IEEE 21st International Symposium on Biomedical Imaging (ISBI) (Proceedings of the IEEE International Symposium on Biomedical Imaging). Institute of Electrical and Electronics Engineers (IEEE). https://doi.org/10.1109/ISBI56570.2024.10635677

Mayo Diaz de Leon, Perla Jazmin ; Cencini, Matteo ; Fatania, Ketan et al. / Deep Image Priors for Magnetic Resonance Fingerprinting with Pretrained Bloch-Consistent Denoising Autoencoders. 2024 IEEE 21st International Symposium on Biomedical Imaging (ISBI). Institute of Electrical and Electronics Engineers (IEEE), 2024. (Proceedings of the IEEE International Symposium on Biomedical Imaging).

@inproceedings{9b9d3e3443c54b418bcc320a313c6e6d,

title = "Deep Image Priors for Magnetic Resonance Fingerprinting with Pretrained Bloch-Consistent Denoising Autoencoders",

abstract = "The estimation of multi-parametric quantitative maps from Magnetic Resonance Fingerprinting (MRF) compressed sampled acquisitions, albeit successful, remains a challenge due to the high underspampling rate and artifacts naturally occuring during image reconstruction. Whilst state-of-the-art DL methods can successfully address the task, to fully exploit their capabilities they often require training on a paired dataset, in an area where ground truth is seldom available. In this work, we propose a method that combines a deep image prior (DIP) module that, without ground truth and in conjunction with a Bloch consistency enforcing autoencoder, can tackle the problem, resulting in a method faster and of equivalent or better accuracy than DIP-MRF.",

keywords = "magnetic resonance fingerprinting, deep learning, deep image priors, quantitative magnetic resonance imaging",

author = "{Mayo Diaz de Leon}, {Perla Jazmin} and Matteo Cencini and Ketan Fatania and Carolin Pirkl and Marion Menzel and Bjoern Menze and Michela Tosetti and Mohammad Golbabaee",

note = "Publisher Copyright: {\textcopyright} 2024 IEEE.; 21st IEEE International Symposium on Biomedical Imaging, ISBI 2024 ; Conference date: 27-05-2024 Through 30-05-2024",

year = "2024",

month = aug,

day = "22",

doi = "10.1109/ISBI56570.2024.10635677",

language = "English",

isbn = "9798350313345",

series = "Proceedings of the IEEE International Symposium on Biomedical Imaging",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

booktitle = "2024 IEEE 21st International Symposium on Biomedical Imaging (ISBI)",

address = "United States",

url = "https://biomedicalimaging.org/2024/",

}

Mayo Diaz de Leon, PJ, Cencini, M, Fatania, K, Pirkl, C, Menzel, M, Menze, B, Tosetti, M & Golbabaee, M 2024, Deep Image Priors for Magnetic Resonance Fingerprinting with Pretrained Bloch-Consistent Denoising Autoencoders. in 2024 IEEE 21st International Symposium on Biomedical Imaging (ISBI). Proceedings of the IEEE International Symposium on Biomedical Imaging, Institute of Electrical and Electronics Engineers (IEEE), 21st IEEE International Symposium on Biomedical Imaging, Athens, Greece, 27/05/24. https://doi.org/10.1109/ISBI56570.2024.10635677

Deep Image Priors for Magnetic Resonance Fingerprinting with Pretrained Bloch-Consistent Denoising Autoencoders. / Mayo Diaz de Leon, Perla Jazmin; Cencini, Matteo; Fatania, Ketan et al.
2024 IEEE 21st International Symposium on Biomedical Imaging (ISBI). Institute of Electrical and Electronics Engineers (IEEE), 2024. (Proceedings of the IEEE International Symposium on Biomedical Imaging).

Research output: Chapter in Book/Report/Conference proceeding › Conference Contribution (Conference Proceeding)

TY - GEN

T1 - Deep Image Priors for Magnetic Resonance Fingerprinting with Pretrained Bloch-Consistent Denoising Autoencoders

AU - Mayo Diaz de Leon, Perla Jazmin

AU - Cencini, Matteo

AU - Fatania, Ketan

AU - Pirkl, Carolin

AU - Menzel, Marion

AU - Menze, Bjoern

AU - Tosetti, Michela

AU - Golbabaee, Mohammad

PY - 2024/8/22

Y1 - 2024/8/22

N2 - The estimation of multi-parametric quantitative maps from Magnetic Resonance Fingerprinting (MRF) compressed sampled acquisitions, albeit successful, remains a challenge due to the high underspampling rate and artifacts naturally occuring during image reconstruction. Whilst state-of-the-art DL methods can successfully address the task, to fully exploit their capabilities they often require training on a paired dataset, in an area where ground truth is seldom available. In this work, we propose a method that combines a deep image prior (DIP) module that, without ground truth and in conjunction with a Bloch consistency enforcing autoencoder, can tackle the problem, resulting in a method faster and of equivalent or better accuracy than DIP-MRF.

AB - The estimation of multi-parametric quantitative maps from Magnetic Resonance Fingerprinting (MRF) compressed sampled acquisitions, albeit successful, remains a challenge due to the high underspampling rate and artifacts naturally occuring during image reconstruction. Whilst state-of-the-art DL methods can successfully address the task, to fully exploit their capabilities they often require training on a paired dataset, in an area where ground truth is seldom available. In this work, we propose a method that combines a deep image prior (DIP) module that, without ground truth and in conjunction with a Bloch consistency enforcing autoencoder, can tackle the problem, resulting in a method faster and of equivalent or better accuracy than DIP-MRF.

KW - magnetic resonance fingerprinting

KW - deep learning

KW - deep image priors

KW - quantitative magnetic resonance imaging

UR - https://biomedicalimaging.org/2024/

U2 - 10.1109/ISBI56570.2024.10635677

DO - 10.1109/ISBI56570.2024.10635677

M3 - Conference Contribution (Conference Proceeding)

SN - 9798350313345

T3 - Proceedings of the IEEE International Symposium on Biomedical Imaging

BT - 2024 IEEE 21st International Symposium on Biomedical Imaging (ISBI)

PB - Institute of Electrical and Electronics Engineers (IEEE)

T2 - 21st IEEE International Symposium on Biomedical Imaging

Y2 - 27 May 2024 through 30 May 2024

ER -

Mayo Diaz de Leon PJ, Cencini M, Fatania K, Pirkl C, Menzel M, Menze B et al. Deep Image Priors for Magnetic Resonance Fingerprinting with Pretrained Bloch-Consistent Denoising Autoencoders. In 2024 IEEE 21st International Symposium on Biomedical Imaging (ISBI). Institute of Electrical and Electronics Engineers (IEEE). 2024. (Proceedings of the IEEE International Symposium on Biomedical Imaging). doi: 10.1109/ISBI56570.2024.10635677

Deep Image Priors for Magnetic Resonance Fingerprinting with Pretrained Bloch-Consistent Denoising Autoencoders

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Handle.net

Other files and links

Fingerprint

HPC (High Performance Computing) and HTC (High Throughput Computing) Facilities

Cite this

Deep Image Priors for Magnetic Resonance Fingerprinting with Pretrained Bloch-Consistent Denoising Autoencoders

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Handle.net

Other files and links

Fingerprint

Equipment

HPC (High Performance Computing) and HTC (High Throughput Computing) Facilities

Cite this