• DocumentCode
    1470062
  • Title

    A Novel Quasi-Linearization Method for CT Image Reconstruction in Scanners With a Multi-Energy Detector System

  • Author

    Ivakhnenko, Vladimir I.

  • Author_Institution
    Reveal Imaging Technol. Inc., Bedford, MA, USA
  • Volume
    57
  • Issue
    2
  • fYear
    2010
  • fDate
    4/1/2010 12:00:00 AM
  • Firstpage
    870
  • Lastpage
    879
  • Abstract
    A novel quasi-linearization method was developed for CT image reconstruction in scanners with a multi-energy detector system. The non-linear equation relating material attenuation coefficients to measured data was reduced to a system of linear equations by means of piece-wise constant approximation of attenuation functions. A material decomposition algorithm was proposed to obtain a physically meaningful solution of this system for the case of N = 2 (dual-energy CT scanners). Image reconstruction methods (the new one and the classical filtered backprojection with beam-hardening correction) were implemented to measured CT data. The new method consistently provides more accurate density results in image reconstruction for the objects containing materials with significantly different attenuation coefficients.
  • Keywords
    computerised tomography; image reconstruction; medical image processing; piecewise constant techniques; position sensitive particle detectors; CT data; CT image reconstruction; X-ray tomography; attenuation coefficients; attenuation functions; computed tomography; image processing; image reconstruction methods; linear equation; material attenuation coefficients; material decomposition algorithm; multienergy detector system; nonlinear equation; numerical methods; piece-wise constant approximation; quasilinearization method; Attenuation measurement; Computed tomography; Computer simulation; Energy measurement; Image processing; Image reconstruction; Nonlinear equations; X-ray detection; X-ray detectors; X-ray imaging; Computed tomography reconstruction; X-ray tomography; computer simulation; image processing; numerical methods;
  • fLanguage
    English
  • Journal_Title
    Nuclear Science, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9499
  • Type

    jour

  • DOI
    10.1109/TNS.2010.2042066
  • Filename
    5446535