PuSH - Publication Server of Helmholtz Zentrum München

Reithmeir, A.* ; Schnabel, J.A. ; Zimmer, V.A.

Learning physics-inspired regularization for medical image registration with hypernetworks.

In: (SPIE Medical Imaging, 2024, San Diego, California). 1000 20th St, Po Box 10, Bellingham, Wa 98227-0010 Usa: SPIE, 2024. (Proc. SPIE ; 12926)
DOI
Open Access Green as soon as Postprint is submitted to ZB.
Medical image registration aims to identify the spatial deformation between images of the same anatomical region and is fundamental to image-based diagnostics and therapy. To date, the majority of the deep learning-based registration methods employ regularizers that enforce global spatial smoothness, e.g., the diffusion regularizer. However, such regularizers are not tailored to the data and might not be capable of reflecting the complex underlying deformation. In contrast, physics-inspired regularizers promote physically plausible deformations. One such regularizer is the linear elastic regularizer, which models the deformation of elastic material. These regularizers are driven by parameters that define the material’s physical properties. For biological tissue, a wide range of estimations of such parameters can be found in the literature, and it remains an open challenge to identify suitable parameter values for successful registration. To overcome this problem and to incorporate physical properties into learning-based registration, we propose to use a hypernetwork that learns the effect of the physical parameters of a physics-inspired regularizer on the resulting spatial deformation field. In particular, we adapt the HyperMorph framework to learn the effect of the two elasticity parameters of the linear elastic regularizer. Our approach enables the efficient discovery of suitable, data-specific physical parameters at test time. To the best of our knowledge, we are the first to use a hypernetwork to learn physics-inspired regularization for medical image registration. We evaluate our approach on 3D intra-patient lung CT images. The results show that the linear elastic regularizer can yield comparable results to the diffusion regularizer in unsupervised learning-based registration while predicting deformations with fewer foldings. With our method, the adaptation of the physical parameters to the data can successfully be performed at test time. Our code is available at https://github.com/annareithmeir/elastic-regularization-hypermorph.
Altmetric
Additional Metrics?
[➜Log in]
Edit extra informations Login
Publication type Article: Conference contribution
Corresponding Author
Keywords Hypernetworks ; Image Registration ; Linear Elastic Regularization; Framework; Head
ISSN (print) / ISBN 0277-786X
e-ISSN 1996-756X
Conference Title SPIE Medical Imaging, 2024
Conference Location San Diego, California
Journal Proceedings of SPIE
Quellenangaben Volume: 12926 Issue: , Pages: , Article Number: , Supplement: ,
Publisher SPIE
Publishing Place 1000 20th St, Po Box 10, Bellingham, Wa 98227-0010 Usa
Non-patent literature Publications
Reviewing status Peer reviewed
Institute(s) Institute for Machine Learning in Biomed Imaging (IML)