Non-manifold mesh extraction from time-varying segmented volumes used for modeling a human heart

  • Authors:
  • Martin Bertram;Gerd Reis;Rolf H. van Lengen;Sascha Köhn;Hans Hagen

  • Affiliations:
  • Center for Artificial Intelligence, Kaiserslautern, Germany;Center for Artificial Intelligence, Kaiserslautern, Germany;Center for Artificial Intelligence, Kaiserslautern, Germany;Center for Artificial Intelligence, Kaiserslautern, Germany;Center for Artificial Intelligence, Kaiserslautern, Germany

  • Venue:
  • EUROVIS'05 Proceedings of the Seventh Joint Eurographics / IEEE VGTC conference on Visualization
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

We present a new algorithm extracting and fairing surfaces from segmented volumes composed of multiple materials. In a first pass, the material boundaries in the volume are smoothed considering signed distance functions for the individual materials. Second, we apply a marching-cubes-like contouring method providing initial meshes defining material boundaries. Non-manifold features emerge along lines where more than two materials encounter. Finally, the mesh geometry is relaxed in a constrained fairing process. We use our algorithm to construct a heart model from segmented time-varying magnetic resonance images. Information concerning the heart ontology is used to merge certain structures to functional units.