A fast and efficient 3D medical image registration method

Mehmet Ali Akinlar, Muhammet Kurulay, Aydin Secer

Research output: Contribution to journalArticlepeer-review

Abstract

A fast and efficient 3D medical image registration method based on adjusting divergence and curl of image displacement field is presented. Using the fact that an image registration problem can be formulated as an optimal control problem, corresponding optimization problem is reduced to solving several Poisson equations. These Poisson equations are solved by finite element (FEM) and multigrid methods (MG), separately. Computational examples indicate that both solution approaches produce similarly good registration quality but that the cost associated with the multigrid approach is, on the average, less than that for the FEM.

Original languageEnglish
Pages (from-to)199-202
Number of pages4
JournalPrzeglad Elektrotechniczny
Volume89
Issue number2 A
Publication statusPublished - 2013
Externally publishedYes

Keywords

  • Finite element
  • Image registration
  • Multigrid method
  • Optimization

Fingerprint

Dive into the research topics of 'A fast and efficient 3D medical image registration method'. Together they form a unique fingerprint.

Cite this