English
 
User Manual Privacy Policy Disclaimer Contact us
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT
  Comparative Quantitative Evaluation of MR-Based Attenuation Correction Methods in Combined Brain PET/MR

Mantlik, F., Hofmann, M., Bezrukov, I., Kolb, A., Beyer, T., Reimold, M., et al. (2010). Comparative Quantitative Evaluation of MR-Based Attenuation Correction Methods in Combined Brain PET/MR. In 2010 Nuclear Science Symposium and Medical Imaging Conference (NSS-MIC 2010).

Item is

Basic

show hide
Item Permalink: http://hdl.handle.net/11858/00-001M-0000-0013-BD9E-9 Version Permalink: http://hdl.handle.net/21.11116/0000-0002-991D-2
Genre: Meeting Abstract

Files

show Files

Locators

show
hide
Locator:
http://ewh.ieee.org/soc/nps/nss-mic/2010/ (Table of contents)
Description:
-

Creators

show
hide
 Creators:
Mantlik, F1, 2, Author              
Hofmann, M1, 2, Author              
Bezrukov, I1, 2, Author              
Kolb , A, Author
Beyer, T, Author
Reimold , M, Author
Pichler, BJ, Author
Schölkopf, B1, 2, Author              
Affiliations:
1Department Empirical Inference, Max Planck Institute for Biological Cybernetics, Max Planck Society, ou_1497795              
2Max Planck Institute for Biological Cybernetics, Max Planck Society, Spemannstrasse 38, 72076 Tübingen, DE, ou_1497794              

Content

show
hide
Free keywords: -
 Abstract: Combined PET/MR provides at the same time molecular and functional imaging as well as excellent soft tissue contrast. It does not allow one to directly measure the attenuation properties of scanned tissues, despite the fact that accurate attenuation maps are necessary for quantitative PET imaging. Several methods have therefore been proposed for MR-based attenuation correction (MR-AC). So far, they have only been evaluated on data acquired from separate MR and PET scanners. We evaluated several MR-AC methods on data from 10 patients acquired on a combined BrainPET/MR scanner. This allowed the consideration of specific PET/MR issues, such as the RF coil that attenuates and scatters 511 keV gammas. We evaluated simple MR thresholding methods as well as atlas and machine learning-based MR-AC. CT-based AC served as gold standard reference. To comprehensively evaluate the MR-AC accuracy, we used RoIs from 2 anatomic brain atlases with different levels of detail. Visual inspection of the PET images indicated that even the basic FLASH threshold MR-AC may be sufficient for several applications. Using a UTE sequence for bone prediction in MR-based thresholding occasionally led to false prediction of bone tissue inside the brain, causing a significant overestimation of PET activity. Although it yielded a lower mean underestimation of activity, it exhibited the highest variance of all methods. The atlas averaging approach had a smaller mean error, but showed high maximum overestimation on the RoIs of the more detailed atlas. The Nave Bayes and Atlas-Patch MR-AC yielded the smallest variance, and the Atlas-Patch also showed the smallest mean error. In conclusion, Atlas-based AC using only MR information on the BrainPET/MR yields a high level of accuracy that is sufficient for clinical quantitative imaging requirements. The Atlas-Patch approach was superior to alternative atlas-based methods, yielding a quantification error below 10 for all RoIs except very small ones.

Details

show
hide
Language(s):
 Dates: 2010-11
 Publication Status: Published in print
 Pages: -
 Publishing info: -
 Table of Contents: -
 Rev. Method: -
 Identifiers: BibTex Citekey: 6801
 Degree: -

Event

show
hide
Title: 2010 Nuclear Science Symposium and Medical Imaging Conference (NSS-MIC 2010)
Place of Event: Knoxville, TN, USA
Start-/End Date: 2010-10-30 - 2010-11-06

Legal Case

show

Project information

show

Source 1

show
hide
Title: 2010 Nuclear Science Symposium and Medical Imaging Conference (NSS-MIC 2010)
Source Genre: Proceedings
 Creator(s):
Affiliations:
Publ. Info: -
Pages: - Volume / Issue: - Sequence Number: M08-4 Start / End Page: - Identifier: -