• DocumentCode
    820808
  • Title

    A regularized deconvolution-fitting method for Compton-scatter correction in SPECT

  • Author

    Wang, Xiaohan ; Koral, Kenneth F.

  • Author_Institution
    Michigan Univ., Ann Arbor, MI, USA
  • Volume
    11
  • Issue
    3
  • fYear
    1992
  • fDate
    9/1/1992 12:00:00 AM
  • Firstpage
    351
  • Lastpage
    360
  • Abstract
    A method is presented for estimating the Compton-scatter component within the photopeak for local energy spectra measured by an Anger camera in SPECT. Assuming that the measured energy spectrum is the source scatter energy distribution convolved with a known camera energy-resolution function plus an unscattered spectral component, a least-square inverse operation is performed to recover the source scatter distribution. Since this inverse operation is ill-posed, the regularization technique is applied for stabilization. With the method, scatter fractions similar to those from polynomial spectral fitting (PSF) have been observed for experimentally measured, high-count data with a hot (Tc99m or I131) sphere in a cold cylinder, and the inverse (Tc99m only). The method is also less sensitive to the width of the fitting window. A regularization parameter from 1 to 10 is recommended for practical cases. The shape of a recovered source scatter distribution matches that determined by a high-resolution semiconductor-detector measurement as well as by Monte Carlo simulation
  • Keywords
    Compton effect; computerised tomography; radioisotope scanning and imaging; 131I; 99mTc; Anger camera; Compton-scatter correction; Monte Carlo simulation; camera energy-resolution function; cold cylinder; fitting window; high-resolution semiconductor-detector measurement; inverse operation; least-square inverse operation; local energy spectra; medical diagnostic imaging; nuclear medicine; photopeak; polynomial spectral fitting; regularization technique; regularized deconvolution-fitting method; single photon emission computerised tomography; source scatter energy distribution; unscattered spectral component; Cameras; Electromagnetic scattering; Energy measurement; Energy resolution; Least squares methods; Nuclear medicine; Particle scattering; Polynomials; Shape measurement; Single photon emission computed tomography;
  • fLanguage
    English
  • Journal_Title
    Medical Imaging, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0278-0062
  • Type

    jour

  • DOI
    10.1109/42.158939
  • Filename
    158939