The aim of this project is to generalize to tree extraction the algorithm that was first developed for building trajectories for virtual endoscopy inside 3D medical images (see 3D Virtual endoscopy page).
First target was to derive efficient potentials for the Fast-Marching algorithm to apply it to the particular problem of tubular anatomical objects.
Second target was to find an intelligent stopping criterion for the front propagation algorithm which speed always remain strictly positive and which cannot integrate any curvature term.
The result is the extraction of trees for typical anatomical datasets like arterial and bronchial trees.
|
|
|
| Tree Extraction in the Abdominal Aorta | Tree Extraction in the Bronchi/Lungs |