We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CV

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computer Vision and Pattern Recognition

Title: NODEO: A Neural Ordinary Differential Equation Based Optimization Framework for Deformable Image Registration

Abstract: Deformable image registration (DIR), aiming to find spatial correspondence between images, is one of the most critical problems in the domain of medical image analysis. In this paper, we present a novel, generic, and accurate diffeomorphic image registration framework that utilizes neural ordinary differential equations (NODEs). We model each voxel as a moving particle and consider the set of all voxels in a 3D image as a high-dimensional dynamical system whose trajectory determines the targeted deformation field. Our method leverages deep neural networks for their expressive power in modeling dynamical systems, and simultaneously optimizes for a dynamical system between the image pairs and the corresponding transformation. Our formulation allows various constraints to be imposed along the transformation to maintain desired regularities. Our experiment results show that our method outperforms the benchmarks under various metrics. Additionally, we demonstrate the feasibility to expand our framework to register multiple image sets using a unified form of transformation,which could possibly serve a wider range of applications.
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Cite as: arXiv:2108.03443 [cs.CV]
  (or arXiv:2108.03443v5 [cs.CV] for this version)

Submission history

From: Yifan Wu [view email]
[v1] Sat, 7 Aug 2021 12:54:17 GMT (10527kb,D)
[v2] Fri, 27 Aug 2021 15:23:18 GMT (21057kb,D)
[v3] Sun, 13 Feb 2022 17:21:14 GMT (58469kb,D)
[v4] Fri, 8 Apr 2022 21:28:04 GMT (4576kb,D)
[v5] Thu, 19 May 2022 04:05:11 GMT (9152kb,D)

Link back to: arXiv, form interface, contact.