• DocumentCode
    307561
  • Title

    Automated detection of the brain intercommissural line from SPECT data

  • Author

    Guimond, AIexandre ; Meunier, Jean ; Soucy, Jean-Paul

  • Author_Institution
    Dept. d´´Inf. et de Recherche Oper., Montreal Univ., Que., Canada
  • Volume
    1
  • fYear
    1995
  • fDate
    20-25 Sep 1995
  • Firstpage
    453
  • Abstract
    The identification of the intercommissural (AC-PC) line is important to achieve registration of nuclear medicine images with a brain atlas. We have adapted to single photon emission computed tomography (SPECT) and optimized a method for automatic detection of this line in positron emission tomography (PET). Briefly, the mid-sagittal plane of the brain is first determined using a similarity criterion between the left and right lobe of the brain. Then, four internal landmarks are detected using diverse image processing and numerical techniques such as edge detection, interpolation and function approximation. Accuracy of the method were previously validated in PET and results in SPECT are in agreement with the PET studies
  • Keywords
    brain; edge detection; function approximation; image registration; interpolation; medical image processing; single photon emission computed tomography; PET; SPECT data; automated detection; brain atlas; brain intercommissural line; diverse image processing; edge detection; function approximation; identification; internal landmarks; interpolation; left lobe; mid-sagittal plane; nuclear medicine image registration; numerical techniques; positron emission tomography; right lobe; similarity criterion; single photon emission computed tomography; Brain; Cost function; Function approximation; Image edge detection; Image processing; Interpolation; Nuclear medicine; Optimization methods; Positron emission tomography; Single photon emission computed tomography;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Engineering in Medicine and Biology Society, 1995., IEEE 17th Annual Conference
  • Conference_Location
    Montreal, Que.
  • Print_ISBN
    0-7803-2475-7
  • Type

    conf

  • DOI
    10.1109/IEMBS.1995.575196
  • Filename
    575196