Cardiac MR Motion Artefact Correction from K-space Using Deep Learning-Based Reconstruction

TY - JOUR

T1 - Cardiac MR Motion Artefact Correction from K-space Using Deep Learning-Based Reconstruction

AU - Oksuz, Ilkay

AU - Clough, James

AU - Bustin, Aurelien

AU - Cruz, Gastao

AU - Prieto, Claudia

AU - Botnar, Rene

AU - Rueckert, Daniel

AU - Schnabel, Julia A.

AU - King, Andrew P.

PY - 2018/9/12

Y1 - 2018/9/12

N2 - Incorrect ECG gating of cardiac magnetic resonance (CMR) acquisitions can lead to artefacts, which hampers the accuracy of diagnostic imaging. Therefore, there is a need for robust reconstruction methods to ensure high image quality. In this paper, we propose a method to automatically correct motion-related artefacts in CMR acquisitions during reconstruction from k-space data. Our method is based on the Automap reconstruction method, which directly reconstructs high quality MR images from k-space using deep learning. Our main methodological contribution is the addition of an adversarial element to this architecture, in which the quality of image reconstruction (the generator) is increased by using a discriminator. We train the reconstruction network to automatically correct for motion-related artefacts using synthetically corrupted CMR k-space data and uncorrupted reconstructed images. Using 25000 images from the UK Biobank dataset we achieve good image quality in the presence of synthetic motion artefacts, but some structural information was lost. We quantitatively compare our method to a standard inverse Fourier reconstruction. In addition, we qualitatively evaluate the proposed technique using k-space data containing real motion artefacts.

AB - Incorrect ECG gating of cardiac magnetic resonance (CMR) acquisitions can lead to artefacts, which hampers the accuracy of diagnostic imaging. Therefore, there is a need for robust reconstruction methods to ensure high image quality. In this paper, we propose a method to automatically correct motion-related artefacts in CMR acquisitions during reconstruction from k-space data. Our method is based on the Automap reconstruction method, which directly reconstructs high quality MR images from k-space using deep learning. Our main methodological contribution is the addition of an adversarial element to this architecture, in which the quality of image reconstruction (the generator) is increased by using a discriminator. We train the reconstruction network to automatically correct for motion-related artefacts using synthetically corrupted CMR k-space data and uncorrupted reconstructed images. Using 25000 images from the UK Biobank dataset we achieve good image quality in the presence of synthetic motion artefacts, but some structural information was lost. We quantitatively compare our method to a standard inverse Fourier reconstruction. In addition, we qualitatively evaluate the proposed technique using k-space data containing real motion artefacts.

KW - Automap

KW - Cardiac MR

KW - Deep learning

KW - Image artefacts

KW - Image quality

KW - Image reconstruction

KW - UK Biobank

UR - http://www.scopus.com/inward/record.url?scp=85053935756&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-00129-2_3

DO - 10.1007/978-3-030-00129-2_3

M3 - Article

AN - SCOPUS:85053935756

SN - 0302-9743

SP - 21

EP - 29

JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

T2 - 1st Workshop on Machine Learning for Medical Image Reconstruction, MLMIR 2018 Held in Conjunction with 21st Medical Image Computing and Computer Assisted Intervention, MICCAI 2018

Y2 - 16 September 2018 through 16 September 2018

ER -