• DocumentCode
    3475021
  • Title

    Experimental results with dual-energy penalized weighted least-squares image reconstruction for X-ray transmission tomography

  • Author

    Sukovic, Predrag ; Clinthorne, Neal H.

  • Author_Institution
    Dept. of Biomed. Eng., Michigan Univ., Ann Arbor, MI, USA
  • Volume
    3
  • fYear
    2000
  • fDate
    2000
  • Abstract
    The authors experimentally evaluate the performance of our previously reported penalized weighted-least squares (PWLS) dual-energy (DE) method. The method is expected to ultimately find its application in positron emission tomography (PET) attenuation correction and quantitative computed tomography (QCT). Both applications require highly accurate density estimates along with a low radiation dose, or in other words low X-ray flux. The authors´ earlier study on computer-generated sets of data showed that De PWLS method is superior to a standard FBP-based DE method. The authors are particularly interested to what extent the performance of their method is affected by photon scatter in the body and the polychromaticity of the X-ray source, as neither scatter nor polychromaticity were taken into account in the computer-generated sets of data. The accuracy of the method and its dependence on X-ray source flux is illustrated and compared to previously reported simulation results
  • Keywords
    computerised tomography; image reconstruction; medical image processing; PET attenuation correction; X-ray source polychromaticity; X-ray transmission tomography; dual-energy penalized weighted least-squares image reconstruction; low X-ray flux; medical diagnostic imaging; photon scatter; quantitative computed tomography; Application software; Attenuation; Computational modeling; Computed tomography; Electromagnetic scattering; Optical computing; Particle scattering; Positron emission tomography; X-ray imaging; X-ray scattering;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Nuclear Science Symposium Conference Record, 2000 IEEE
  • Conference_Location
    Lyon
  • ISSN
    1082-3654
  • Print_ISBN
    0-7803-6503-8
  • Type

    conf

  • DOI
    10.1109/NSSMIC.2000.949389
  • Filename
    949389