English
 
Help Privacy Policy Disclaimer
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT
  Evaluation of Deformable Registration Methods for MR-CT Atlas Alignment

Scheel, V., Hofmann, M., Rehfeld, N., Judenhofer, M., Claussen, C., & Pichler, B. (2007). Evaluation of Deformable Registration Methods for MR-CT Atlas Alignment. Poster presented at IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS-MIC 2007), Honolulu, HI, USA.

Item is

Files

show Files

Locators

show
hide
Description:
-

Creators

show
hide
 Creators:
Scheel, V, Author
Hofmann, M1, Author              
Rehfeld, N, Author
Judenhofer , MS, Author
Claussen, CD, Author
Pichler, BJ, Author
Affiliations:
1Department Empirical Inference, Max Planck Institute for Biological Cybernetics, Max Planck Society, ou_1497795              

Content

show
hide
Free keywords: -
 Abstract: Deformable registration methods are essential for multimodality imaging. Many different methods exist but due to the complexity of the deformed images a direct comparison of the methods is difficult. One particular application that requires high accuracy registration of MR-CT images is atlas-based attenuation correction for PET/MR. We compare four deformable registration algorithms for 3D image data included in the Open Source "National Library of Medicine Insight Segmentation and Registration Toolkit" (ITK). An interactive landmark based registration using MiraView (Siemens) has been used as gold standard. The automatic algorithms provided by ITK are based on the metrics Mattes mutual information as well as on normalized mutual information. The transformations are calculated by interpolating over a uniform B-Spline grid laying over the image to be warped. The algorithms were tested on head images from 10 subjects. We implemented a measure which segments head interior bone and air based on the CT images and l ow intensity classes of corresponding MRI images. The segmentation of bone is performed by individually calculating the lowest Hounsfield unit threshold for each CT image. The compromise is made by quantifying the number of overlapping voxels of the remaining structures. We show that the algorithms provided by ITK achieve similar or better accuracy than the time-consuming interactive landmark based registration. Thus, ITK provides an ideal platform to generate accurately fused datasets from different modalities, required for example for building training datasets for Atlas-based attenuation correction.

Details

show
hide
Language(s):
 Dates: 2007-11
 Publication Status: Published in print
 Pages: -
 Publishing info: -
 Table of Contents: -
 Rev. Type: -
 Identifiers: BibTex Citekey: 4850
 Degree: -

Event

show
hide
Title: IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS-MIC 2007)
Place of Event: Honolulu, HI, USA
Start-/End Date: 2007-10-27 - 2007-11-03

Legal Case

show

Project information

show

Source 1

show
hide
Title: IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS-MIC 2007)
Source Genre: Proceedings
 Creator(s):
Affiliations:
Publ. Info: -
Pages: - Volume / Issue: - Sequence Number: M13-121 Start / End Page: - Identifier: -