• DocumentCode
    1543506
  • Title

    The computational performance of a high-order coupled FEM/BEM procedure in electropotential problems

  • Author

    Bradley, Chris P. ; Harris, Glen M. ; Pullan, Andrew J.

  • Author_Institution
    Lab. of Physiol., Oxford Univ., UK
  • Volume
    48
  • Issue
    11
  • fYear
    2001
  • Firstpage
    1238
  • Lastpage
    1250
  • Abstract
    Presents a thorough analysis of the computational performance of a coupled cubic Hermite boundary element/finite element procedure. This C 1 (i.e., value and derivative continuous) method has been developed specifically for electropotential problems, and has been previously applied to torso and skull problems. Here, the behavior of this new procedure is quantified by solving a number of dipole in spheres problems. A detailed set of results generated with a wide range of the various input parameters (such as dipole orientation, location, conductivity, and solution method used in each spherical shell [either finite element or boundary elements]) is presented. The new cubic Hermite boundary element procedure shows significantly better accuracy and convergence properties and a significant reduction in CPU time than a traditional boundary element procedure which uses linear or constant elements. Results using the high-order method are also compared with other computational methods which have had quantitative results published for electropotential problems. In all eases, the high-order method offered a significant improvement in computational efficiency by increasing the solution accuracy for the same, or fewer, solution degrees of freedom.
  • Keywords
    bioelectric potentials; boundary-elements methods; finite element analysis; physiological models; computational efficiency; computational performance; conductivity; coupled cubic Hermite boundary element/finite element procedure; derivative continuous method; dipole location; dipole orientation; electropotential problems; high-order coupled FEM/BEM procedure; Biological system modeling; Boundary element methods; Finite element methods; High performance computing; Mathematical model; Numerical models; Performance analysis; Skull; Testing; Torso; Biomedical Engineering; Computer Simulation; Electrocardiography; Electroencephalography; Electrophysiology; Finite Element Analysis; Humans; Models, Biological;
  • fLanguage
    English
  • Journal_Title
    Biomedical Engineering, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9294
  • Type

    jour

  • DOI
    10.1109/10.959319
  • Filename
    959319