Log In Sign Up

Exploring Deep Registration Latent Spaces

by   Théo Estienne, et al.

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.


page 6

page 7

page 12

page 13


Exploring Intensity Invariance in Deep Neural Networks for Brain Image Registration

Image registration is a widely-used technique in analysing large scale d...

Learning Deformable Registration of Medical Images with Anatomical Constraints

Deformable image registration is a fundamental problem in the field of m...

mlVIRNET: Multilevel Variational Image Registration Network

We present a novel multilevel approach for deep learning based image reg...

Hybrid Atlas Building with Deep Registration Priors

Registration-based atlas building often poses computational challenges i...

Train smarter, not harder: learning deep abdominal CT registration on scarce data

Purpose: This study aims to explore training strategies to improve convo...

A Deep-Discrete Learning Framework for Spherical Surface Registration

Cortical surface registration is a fundamental tool for neuroimaging ana...