Physics-informed brain MRI segmentation

TY - CHAP

T1 - Physics-informed brain MRI segmentation

AU - Borges, Pedro

AU - Sudre, Carole

AU - Varsavsky, Thomas

AU - Thomas, David

AU - Drobnjak, Ivana

AU - Ourselin, Sebastien

AU - Cardoso, M. Jorge

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Magnetic Resonance Imaging (MRI) is one of the most flexible and powerful medical imaging modalities. This flexibility does however come at a cost; MRI images acquired at different sites and with different parameters exhibit significant differences in contrast and tissue appearance, resulting in downstream issues when quantifying brain anatomy or the presence of pathology. In this work, we propose to combine multiparametric MRI-based static-equation sequence simulations with segmentation convolutional neural networks (CNN), to make these networks robust to variations in acquisition parameters. Results demonstrate that, when given both the image and their associated physics acquisition parameters, CNNs can produce segmentations that exhibit robustness to acquisition variations. We also show that the proposed physics-informed methods can be used to bridge multi-centre and longitudinal imaging studies where imaging acquisition varies across a site or in time.

AB - Magnetic Resonance Imaging (MRI) is one of the most flexible and powerful medical imaging modalities. This flexibility does however come at a cost; MRI images acquired at different sites and with different parameters exhibit significant differences in contrast and tissue appearance, resulting in downstream issues when quantifying brain anatomy or the presence of pathology. In this work, we propose to combine multiparametric MRI-based static-equation sequence simulations with segmentation convolutional neural networks (CNN), to make these networks robust to variations in acquisition parameters. Results demonstrate that, when given both the image and their associated physics acquisition parameters, CNNs can produce segmentations that exhibit robustness to acquisition variations. We also show that the proposed physics-informed methods can be used to bridge multi-centre and longitudinal imaging studies where imaging acquisition varies across a site or in time.

KW - Deep learning

KW - Harmonization

KW - MRI

UR - http://www.scopus.com/inward/record.url?scp=85075647010&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-32778-1_11

DO - 10.1007/978-3-030-32778-1_11

M3 - Conference paper

AN - SCOPUS:85075647010

SN - 9783030327774

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 100

EP - 109

BT - Simulation and Synthesis in Medical Imaging - 4th International Workshop, SASHIMI 2019, Held in Conjunction with MICCAI 2019, Proceedings

A2 - Burgos, Ninon

A2 - Gooya, Ali

A2 - Svoboda, David

PB - SPRINGER

T2 - 4th International Workshop on Simulation and Synthesis in Medical Imaging, SASHIMI 2019, held in conjunction with the 22nd International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2019

Y2 - 13 October 2019 through 13 October 2019

ER -