mlVIRNET: Multilevel Variational Image Registration Network

A. Hering, B. van Ginneken and S. Heldmann

in: Medical Image Computing and Computer-Assisted Intervention, volume 11769 of Lecture Notes in Computer Science, 2019, pages 724-731

DOI arXiv

Abstract

We present a novel multilevel approach for deep learning based image registration. Recently published deep learning based registration methods have shown promising results for a wide range of tasks. However, these algorithms are still limited to relatively small deformations. Our method addresses this shortcoming by introducing a multilevel framework, which computes deformation fields on different scales, similar to conventional methods. Thereby, a coarse-level alignment is obtained first, which is subsequently improved on finer levels. We demonstrate our method on the complex task of inhale-to-exhale lung registration. We show that the use of a deep learning multilevel approach leads to significantly better registration results.

A pdf file of this publication is available for personal use. Enter your e-mail address in the box below and press the button. You will receive an e-mail message with a link to the pdf file.