• DocumentCode
    1305372
  • Title

    Coherent Structures of Characteristic Curves in Symmetric Second Order Tensor Fields

  • Author

    Hlawatsch, Marcel ; Vollrath, Joachim E. ; Sadlo, Filip ; Weiskopf, Daniel

  • Author_Institution
    Visualization Res. Center, Univ. Stuttgart (VISUS), Stuttgart, Germany
  • Volume
    17
  • Issue
    6
  • fYear
    2011
  • fDate
    6/1/2011 12:00:00 AM
  • Firstpage
    781
  • Lastpage
    794
  • Abstract
    This paper generalizes the concept of Lagrangian coherent structures, which is known for its potential to visualize coherent regions in vector fields and to distinguish them from each other. In particular, we extend the concept of the flow map to generic mappings of coordinates. As the major application of this generalization, we present a semiglobal method for visualizing coherent structures in symmetric second order tensor fields. We demonstrate the usefulness by examples from DT-MRI, uncovering anatomical structures in linearly anisotropic regions not amenable to local feature criteria. To further exemplify the suitability of our concept, we also present its application to stress tensor fields. Last, an accelerated implementation utilizing GPUs is presented.
  • Keywords
    biomedical MRI; computational geometry; data visualisation; tensors; DT-MRI; GPU; Lagrangian coherent structures; anatomical structures; characteristic curves; coherent structure visualization; diffusion tensor magnetic resonance imaging; flow map; linearly anisotropic regions; stress tensor fields; symmetric second order tensor fields; tensor field visualization; Anisotropic magnetoresistance; Diffusion tensor imaging; Eigenvalues and eigenfunctions; Feature extraction; Tensile stress; Visualization; Lagrangian coherent structures; feature extraction; general-purpose computation using graphics hardware.; tensor field topology; tensor field visualization;
  • fLanguage
    English
  • Journal_Title
    Visualization and Computer Graphics, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    1077-2626
  • Type

    jour

  • DOI
    10.1109/TVCG.2010.107
  • Filename
    5557866