• DocumentCode
    3342033
  • Title

    Incorporation of a cascade gamma ray correction into the SRW iterative reconstruction for non-standard PET nuclides: Towards a unified correction weighted (UCW) scheme in the sensitivity image

  • Author

    Cheng, Ju-Chieh ; Laforest, Richard

  • Author_Institution
    Sch. of Med., Dept. of Radiol., Washington Univ., St. Louis, MO, USA
  • fYear
    2011
  • fDate
    23-29 Oct. 2011
  • Firstpage
    4059
  • Lastpage
    4062
  • Abstract
    We describe an extension of the scatter and randoms weighted (SRW) iterative PET reconstruction algorithm which includes a model-based cascade estimate in the trues fraction (TF) estimation for non-standard PET nuclides as an effort to unify the corrections for quantitative PET imaging into the sensitivity image. The unified correction weighted (UCW) scheme which includes the attenuation, normalization, and TF estimates was incorporated into the widely used 3D-OSEM reconstruction. A 76Br contrast phantom with a ~54% background contamination (including the scatter, randoms, and cascade events) and a 86Y contrast phantom with a ~83% background contamination were scanned, and the data were reconstructed using the standard and the UCW methods. The contrast vs noise and the reconstruction time comparisons were evaluated, and it was found that the UCW method outperforms the standard Ordinary Poisson (OP) approach in terms of convergence in hot and cold contrasts. A 50% overall time gain for the 76Br study and an 80% time gain for the 86Y study as compared to the OP method was achieved using the proposed method due to the better convergence rate (i.e. able to stop at earlier iteration) and faster reconstruction time (i.e. no need to store and process the scatter, randoms, and cascade sinograms in the forward model for every single iteration). These studies also demonstrated that the background correction weighted scheme in the sensitivity image works well with relatively high background fractions. It was observed that the higher the background contamination, the better the convergence rate for the UCW method as compared to the OP approach.
  • Keywords
    expectation-maximisation algorithm; image reconstruction; medical image processing; phantoms; positron emission tomography; stochastic processes; 3D-OSEM reconstruction; SRW iterative reconstruction; UCW method; background correction weighted scheme; cascade gamma ray correction; contrast phantom; image reconstruction; non-standard PET nuclides; ordinary Poisson approach; quantitative PET imaging; reconstruction time; sensitivity image; true fraction estimation; unified correction weighted scheme; Noise;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC), 2011 IEEE
  • Conference_Location
    Valencia
  • ISSN
    1082-3654
  • Print_ISBN
    978-1-4673-0118-3
  • Type

    conf

  • DOI
    10.1109/NSSMIC.2011.6153771
  • Filename
    6153771