Top Left. Semi-automatic segmentation techniques use image gradients to
trace the boundary of the prostate. A series of axial images are segmented.
Every 3-5 images are segmented and used to generate a 3d surface
representation of the anatomy.
Top Right. A volumetric FE mesh is generated by intersecting a structured
grid with the 3D surface as the boundary. The unstructured grid is generated
by pulling, smoothing, and pillowing the subset unstructured hexahedral
elements to conform to the boundary of the anatomy.
Bottom Left. The resulting hexahedral mesh is shown registered to a
volume rendering of the canine anatomy in the prostate region.
Bottom Right. A schematic of the Imaging-to-FEM mesh
pipeline is shown. The hexahedral mesh shown has distinct regions for
the prostate and laser applicator.