Image processing and grid generation for finite element analysis of defibrillation

To generate a realistic finite element model for electrical defibrillation studies, torso images are used to provide geometry and conductivity information. For uniform grids, images have been segmented into different tissue types in the past in order to provide material type information for the rectangular elements. For nonuniform grids, boundaries of anatomical structures were extracted manually for grid generation, which was a labor-intensive process. In this paper, we present stack image processing techniques and a polygonal boundary approximation algorithm which will allow more automatic and accurate generation of both uniform and nonuniform grids.