The Meshing Module

The goal here is to generate a 3D mesh of the blood vessels visible in the MRI. In order to reach this goal, the MRI must undergo several processing steps which are detailed below.

1. Segmentation & Surface extraction (VMTK)

Image segmentation is used to detect blood vessels in the MRIs. Blood vessels surfaces are extracted (using a level-set method). See Surface extraction for details.

2. Center lines definition (VTK)

Blood vessels center lines are defined and can be corrected by hand using a graphical tool. For each center line point, the inscribed sphere is also stored.

For more informations, refer to The centerlines editing tool.

3. Center lines processing (VMTK,VTK)

A 3D-mesh is created, defining the center lines. Smoothing of the mesh is performed using splines. See Center lines computing for more informations.

4. Image reconstruction (VMTK,VTK)

Using the center lines and a computed signed distance function, we generate a high quality, artifact-free image with controlled precision. See Image generation for more informations.

5. Surface extraction (VMTK)

Here we extract the blood vessels surfaces based on the zero level of the distance function. This process is the same we used in the first step (Segmentation & surface extraction), except we work on the 3D image generated using the center lines only, which is cleaner and smoother. See Surface extraction for more details.

6. Numerical simulation mesh construction (GMSH)

The mesh is prepared to be used in computer simulation:

6.1. Part 1: Surface processing

6.2. Part 2: Volume processing

  • The volume is meshed using the center lines.

  • Optionnally, the blood vessels thickness is taken into account by extruding de boundary layer according to a given percentage of the blood vessels radius.

  • The Mesh is partitioned for numerical simulation. See Volume meshing for more details.