Brain-cloud: A generalized and flexible registration framework for brain MR images

Minjeong Kim, Guorong Wu, Qian Wang, Dinggang Shen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Image registration, which aligns a pair of fixed and moving images, is often tackled by the large shape and intensity variation between the images. As a remedy, we present a generalized registration framework that is capable to predict the initial deformation field between the fixed and moving images, even though their appearances are very different. For the prediction, we learn the prior knowledge on deformation from pre-observed images. Especially, our method is significantly differentiated from previous methods that are usually confined to a specific fixed image, to be flexible for handling arbitrary fixed and moving images. Specifically, our idea is to encapsulate many pre-observed images into a hierarchical infrastructure, termed as cloud, which is able to efficiently compute the deformation pathways between the pre-observed images. After anchoring the fixed and moving images to their respective port images (similar images in terms of intensity appearance) in the cloud, we predict the initial deformation between the fixed and moving images by the deformation pathway between the two port images. Thus, the remaining small deformation can be efficiently refined via most existing deformable registration methods. With the cloud, we have obtained promising registration results on both adult and infant brain images, demonstrating the advantage of the proposed registration framework in improving the registration performance.

Original languageEnglish
Title of host publicationAugmented Reality Environments for Medical Imaging and Computer-Assisted Interventions - 6th Int. Workshop, MIAR 2013 and 8th Int. Workshop, AE-CAI 2013, Held in Conjunction with MICCAI 2013, Proc.
Pages153-161
Number of pages9
DOIs
Publication statusPublished - 2013
Event6th International Workshop on Augmented Reality Environments for Medical Imaging and Computer-Assisted Interventions, MIAR 2013 and 8th International Workshop, AE-CAI 2013, Held in Conjunction with MICCAI 2013 - Nagoya, Japan
Duration: 2013 Sep 222013 Sep 22

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8090 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other6th International Workshop on Augmented Reality Environments for Medical Imaging and Computer-Assisted Interventions, MIAR 2013 and 8th International Workshop, AE-CAI 2013, Held in Conjunction with MICCAI 2013
CountryJapan
CityNagoya
Period13/9/2213/9/22

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Brain-cloud: A generalized and flexible registration framework for brain MR images'. Together they form a unique fingerprint.

  • Cite this

    Kim, M., Wu, G., Wang, Q., & Shen, D. (2013). Brain-cloud: A generalized and flexible registration framework for brain MR images. In Augmented Reality Environments for Medical Imaging and Computer-Assisted Interventions - 6th Int. Workshop, MIAR 2013 and 8th Int. Workshop, AE-CAI 2013, Held in Conjunction with MICCAI 2013, Proc. (pp. 153-161). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8090 LNCS). https://doi.org/10.1007/978-3-642-40843-4_17