• DocumentCode
    3173998
  • Title

    A new algorithm for the attenuation correction in SPECT

  • Author

    Voegelin, M.R. ; Jafrancesco, D. ; Paoli, G.

  • Author_Institution
    Sect. of Phys., Dept. of Pathophysiol., Italy
  • fYear
    1995
  • fDate
    10-13 Sept. 1995
  • Firstpage
    525
  • Lastpage
    528
  • Abstract
    The proposed method is based on the reconstruction of the activity distribution c(ij) on a section S, with ij=1 to 64, assuming that the attenuation coefficients /spl mu/(ij) on the section are known. This method, for every section, stores in a matrix 64/spl times/64 elements N projections of the section acquired at angles k/spl phi//sub 0/(k-1 to N, /spl phi//sub 0/=360/sup c//N). The simplified hypothesis that along k the radioactive concentration c changes in a linear way is made. Using a best fit method the values of radioactive concentration along this direction are calculated and the correct projection is obtained. The algorithm of filtered back-projection is used to reconstruct the distribution c(ij) on the section. The method was tested using theoretical phantoms of different shapes, locations and radioactive types. The method needs very short times because the attenuation correction and the reconstruction procedure are performed at the same time.
  • Keywords
    gamma-ray absorption; image reconstruction; medical image processing; single photon emission computed tomography; SPECT; activity distribution; algorithm; attenuation coefficients; attenuation correction; best fit method; filtered back-projection; matrix; phantoms; radioactive concentration; radioactive types; reconstruction; Attenuation; Cameras; Cardiology; Distributed computing; Imaging phantoms; Optical attenuators; Optical filters; Physics; Shape; Testing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computers in Cardiology 1995
  • Conference_Location
    Vienna, Austria
  • Print_ISBN
    0-7803-3053-6
  • Type

    conf

  • DOI
    10.1109/CIC.1995.482717
  • Filename
    482717