• DocumentCode
    2433803
  • Title

    The watershed transformation on graphs for the segmentation of CT images

  • Author

    Wegner, S. ; Harms, T. ; Oswald, H. ; Fleck, E.

  • Author_Institution
    Humboldt-Univ., Berlin, Germany
  • Volume
    3
  • fYear
    1996
  • fDate
    25-29 Aug 1996
  • Firstpage
    498
  • Abstract
    A segmentation method for computer tomography images using the concept of the watershed transformation on graphs in a hyperthermia planning environment is presented. First, the watershed transformation is applied on the gradient image. The resulting mosaic image is transformed into a graph and the watershed transformation is again performed on this graph. Solutions for the difficulties on graphs are presented. The iteration of this procedure leads to multiresolution segmentation planes showing regions that describe anatomical objects of different resolution at each plane. For some objects the selection of the adequate plane can be based on an object description. Since this is not possible for all objects, the segmentation planes are used as basis for an interactive segmentation environment. The user can select objects in different planes or merge regions to obtain an object
  • Keywords
    computerised tomography; graph theory; hyperthermia; image segmentation; interactive systems; iterative methods; medical image processing; microwave heating; radiation therapy; 70 to 90 MHz; CT image segmentation; EM radiation heating; anatomical objects; computer tomography; gradient image; graphs; hyperthermia planning environment; interactive segmentation environment; iteration; mosaic image; multiresolution segmentation planes; watershed transformation; Biomedical imaging; Computational modeling; Computed tomography; Electromagnetic radiation; Heart; Hyperthermia; Image segmentation; Medical treatment; Neoplasms; Temperature distribution;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Pattern Recognition, 1996., Proceedings of the 13th International Conference on
  • Conference_Location
    Vienna
  • ISSN
    1051-4651
  • Print_ISBN
    0-8186-7282-X
  • Type

    conf

  • DOI
    10.1109/ICPR.1996.546997
  • Filename
    546997