Skip to main navigation Skip to search Skip to main content

Groupwise registration of brain images for establishing accurate spatial correspondence of brain structures

  • Zhenyu Tang
  • , Yong Fan*
  • *Corresponding author for this work
  • CAS - Institute of Automation

Research output: Contribution to journalArticlepeer-review

Abstract

For establishing accurate spatial correspondence of brain structures among different subjects, many groupwise image registrationmethods have been proposed to register brain images taken from different subjects onto a common space. Except the congealing method, most groupwise image registration methods achieve the image registration by registering images to a template image using pairwise image registration algorithms. For these groupwise image registration methods built upon pairwise image registration, the key points are template determination, registration path identification, and pairwise image registration. Focusing on the graph-based groupwise image registration methods due to their high computation efficiency and accuracy, this chapter introduces briefly the congealing method and groupwise image registration methods with different strategies for template determination and registration path identification. To demonstrate the strength of state-of-the-art groupwise image registrationmethods, a quantitative comparison study has also been presented for representative graph-based groupwise image registrationmethods based on two publicly available 3D MR brain image datasets.

Original languageEnglish
Pages (from-to)229
Number of pages1
JournalLecture Notes in Computational Vision and Biomechanics
Volume14
DOIs
StatePublished - 2014
Externally publishedYes

Fingerprint

Dive into the research topics of 'Groupwise registration of brain images for establishing accurate spatial correspondence of brain structures'. Together they form a unique fingerprint.

Cite this