• DocumentCode
    2029443
  • Title

    Emerging segmentation of neuroanatomical structures in magnetic resonance images

  • Author

    Rajapakse, Jagath C. ; Tan, Choong Leong

  • Author_Institution
    Sch. of Appl. Sci., Nanyang Technol. Univ., Singapore
  • Volume
    3
  • fYear
    1999
  • fDate
    1999
  • Firstpage
    910
  • Abstract
    The emerging scheme (J.C. Rajapakse and F. Kruggel) is used to segment neuroanatomical structures in magnetic resonance (MR) head scans. Emerging is a region growing technique which is guided by the edgeness profile of the image. It is similar to the watershed algorithm (L. Vincent and P. Soille, 1991), but allows splitting of regions in catchment basins and merging at watershed boundaries. Region boundaries in MR head scans are fuzzy and compositions of tissues of some structures render varying intensity patterns. These characteristics and scanner limitations pose difficulties to unsupervised segmentation of MR head scans. We present emerging as a segmentation technique that overcomes these difficulties to produce reasonable segmentations of neuroanatomical structures
  • Keywords
    biomedical MRI; brain; image segmentation; medical image processing; neurophysiology; catchment basins; edgeness profile; emerging scheme; emerging segmentation; magnetic resonance head scans; magnetic resonance images; neuroanatomical structure segmentation; region boundaries; region growing technique; region splitting; scanner limitations; tissue compositions; unsupervised segmentation; varying intensity patterns; watershed algorithm; Brain; Humans; Image segmentation; Magnetic fields; Magnetic heads; Magnetic resonance; Magnetic resonance imaging; Parkinson´s disease; Rendering (computer graphics); Shape measurement;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Neural Information Processing, 1999. Proceedings. ICONIP '99. 6th International Conference on
  • Conference_Location
    Perth, WA
  • Print_ISBN
    0-7803-5871-6
  • Type

    conf

  • DOI
    10.1109/ICONIP.1999.844658
  • Filename
    844658