TY - GEN
T1 - Exploring Deep Registration Latent Spaces
AU - Estienne, Théo
AU - Vakalopoulou, Maria
AU - Christodoulidis, Stergios
AU - Battistella, Enzo
AU - Henry, Théophraste
AU - Lerousseau, Marvin
AU - Leroy, Amaury
AU - Chassagnon, Guillaume
AU - Revel, Marie Pierre
AU - Paragios, Nikos
AU - Deutsch, Eric
N1 - Publisher Copyright:
© 2021, Springer Nature Switzerland AG.
PY - 2021/1/1
Y1 - 2021/1/1
N2 - Explainability of deep neural networks is one of the most challenging and interesting problems in the field. In this study, we investigate the topic focusing on the interpretability of deep learning-based registration methods. In particular, with the appropriate model architecture and using a simple linear projection, we decompose the encoding space, generating a new basis, and we empirically show that this basis captures various decomposed anatomically aware geometrical transformations. We perform experiments using two different datasets focusing on lungs and hippocampus MRI. We show that such an approach can decompose the highly convoluted latent spaces of registration pipelines in an orthogonal space with several interesting properties. We hope that this work could shed some light on a better understanding of deep learning-based registration methods.
AB - Explainability of deep neural networks is one of the most challenging and interesting problems in the field. In this study, we investigate the topic focusing on the interpretability of deep learning-based registration methods. In particular, with the appropriate model architecture and using a simple linear projection, we decompose the encoding space, generating a new basis, and we empirically show that this basis captures various decomposed anatomically aware geometrical transformations. We perform experiments using two different datasets focusing on lungs and hippocampus MRI. We show that such an approach can decompose the highly convoluted latent spaces of registration pipelines in an orthogonal space with several interesting properties. We hope that this work could shed some light on a better understanding of deep learning-based registration methods.
KW - Deep learning-based medical image registration
KW - Deformable registration
KW - Explainability
UR - http://www.scopus.com/inward/record.url?scp=85116426925&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-87722-4_11
DO - 10.1007/978-3-030-87722-4_11
M3 - Conference contribution
AN - SCOPUS:85116426925
SN - 9783030877217
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 112
EP - 122
BT - Domain Adaptation and Representation Transfer, and Affordable Healthcare and AI for Resource Diverse Global Health - 3rd MICCAI Workshop, DART 2021, and 1st MICCAI Workshop, FAIR 2021, Held in Conjunction with MICCAI 2021, Proceedings
A2 - Albarqouni, Shadi
A2 - Cardoso, M. Jorge
A2 - Dou, Qi
A2 - Kamnitsas, Konstantinos
A2 - Khanal, Bishesh
A2 - Rekik, Islem
A2 - Rieke, Nicola
A2 - Sheet, Debdoot
A2 - Tsaftaris, Sotirios
A2 - Xu, Daguang
A2 - Xu, Ziyue
PB - Springer Science and Business Media Deutschland GmbH
T2 - 3rd MICCAI Workshop on Domain Adaptation and Representation Transfer, DART 2021, and the 1st MICCAI Workshop on Affordable Healthcare and AI for Resource Diverse Global Health, FAIR 2021, held in conjunction with 24th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2021
Y2 - 27 September 2021 through 1 October 2021
ER -