The effect of scatter correction on dual energy micro-CT


Conference Paper

Dual energy (DE) CT imaging is expected to play a major role in the diagnostic arena as it provides a quantitative decomposition of basis materials, opening the door for new clinical applications without significantly increasing dose to the patient. DE-CT provides a particularly unique opportunity in preclinical CT where new elemental contrast agents are providing novel approaches for quantitative tissue characterization. We have implemented DE-CT imaging with a preclinical dual source micro-CT scanner. With this configuration, both forward and cross-scatter can substantially degrade image quality. This work investigated the effect of scatter correction on the accuracy of post-reconstruction iodine and calcium decomposition. Scatter has been estimated using a lead beam stop technique. Our approach involves noise reduction in the scatter corrected images using bilateral filtering. The scatter correction has been quantitatively evaluated using phantom experiments and in vivo cancer imaging. As shown by our measurements, the dual source scanning is affected more by the cross-scatter from the high energy to the low energy imaging chain. The scatter correction reduced the presence of cupping artifacts and increased both the accuracy and precision of dual energy decompositions of calcium and iodine. On average, the root mean square errors in retrieving true iodine and calcium concentrations via dual energy were reduced by 32%. As a result of scatter corrections, we expect more accurate quantification of important vascular biomarkers such as fractional blood volume and vascular permeability in preclinical cancer studies. © 2013 SPIE.

Full Text

Duke Authors

Cited Authors

  • Clark, D; Johnston, SM; Johnson, GA; Badea, CT

Published Date

  • June 3, 2013

Published In

Volume / Issue

  • 8668 /

International Standard Serial Number (ISSN)

  • 1605-7422

International Standard Book Number 13 (ISBN-13)

  • 9780819494429

Digital Object Identifier (DOI)

  • 10.1117/12.2006904

Citation Source

  • Scopus