APSIPA Transactions on Signal and Information Processing > Vol 11 > Issue 1

Self-Supervised Motion-Corrected Image Reconstruction Network for 4D Magnetic Resonance Imaging of the Body Trunk

Thomas Küstner, University Hospital Tübingen, Germany, thomas.kuestner@med.uni-tuebingen.de , Jiazhen Pan, Technical University of Munich, Christopher Gilliam, RMIT University, Haikun Qi, ShanghaiTech University, Gastao Cruz, King’s College London, Kerstin Hammernik, Technical University of Munich and Imperial College London, Thierry Blu, Chinese University Hong Kong, Daniel Rueckert, Technical University of Munich and Imperial College London, René Botnar, King’s College London and Pontificia Universidad Católica de Chile, Claudia Prieto, King’s College London and Pontificia Universidad Católica de Chile, Sergios Gatidis, University Hospital Tübingen
 
Suggested Citation
Thomas Küstner, Jiazhen Pan, Christopher Gilliam, Haikun Qi, Gastao Cruz, Kerstin Hammernik, Thierry Blu, Daniel Rueckert, René Botnar, Claudia Prieto and Sergios Gatidis (2022), "Self-Supervised Motion-Corrected Image Reconstruction Network for 4D Magnetic Resonance Imaging of the Body Trunk", APSIPA Transactions on Signal and Information Processing: Vol. 11: No. 1, e12. http://dx.doi.org/10.1561/116.00000039

Publication Date: 09 May 2022
© 2022 T. Küstner et al.
 
Subjects
 
Keywords
Motion-compensated image reconstructionMagnetic Resonance ImagingImage registrationDeep learning reconstruction
 

Share

Open Access

This is published under the terms of CC BY-NC.

Downloaded: 912 times

In this article:
Introduction 
In-Vivo 4D MR Acquisition 
Non-Rigid Registration in k-Space 
Motion-Corrected Image Reconstruction 
Evaluation and Experiments 
Results 
Discussion 
Conclusion 
References 

Abstract

Respiratory motion can cause artifacts in magnetic resonance imaging of the body trunk if patients cannot hold their breath or triggered acquisitions are not practical. Retrospective correction strategies usually cope with motion by fast imaging sequences under free-movement conditions followed by motion binning based on motion traces. These acquisitions yield sub-Nyquist sampled and motion-resolved k-space data. Motion states are linked to each other by non-rigid deformation fields. Usually, motion registration is formulated in image space which can however be impaired by aliasing artifacts or by estimation from low-resolution images. Subsequently, any motion-corrected reconstruction can be biased by errors in the deformation fields. In this work, we propose a deep-learning based motion-corrected 4D (3D spatial + time) image reconstruction which combines a non-rigid registration network and a 4D reconstruction network. Non-rigid motion is estimated in k-space and incorporated into the reconstruction network. The proposed method is evaluated on in-vivo 4D motion-resolved magnetic resonance images of patients with suspected liver or lung metastases and healthy subjects. The proposed approach provides 4D motion-corrected images and deformation fields. It enables a ∼14× accelerated acquisition with a 25-fold faster reconstruction than comparable approaches under consistent preservation of image quality for changing patients and motion patterns.

DOI:10.1561/116.00000039