[Gate-users] Overlap between scanner geometry and Voxellized phantom

[Niklas Rehfeld]

Hi Bosky,

All that Simon just said is totally right. Geant4 assumes that there is no
overlap between the volumes. That means if there is an overlap you get
probably wrong results. As Simon pointed out, the particle stays in the
volume until it crosses the border of the volume and then looks for the
next volume. In your case that means that the photon stays in Air.

Unfortunately, it is (at present) not possible in Geant4 to do boolean
operations with two volumes if one volume is a parameterized one (which
would be an elegant but not necessarily fast solution to the problem).

To my knowledge, the only thing you can do is to split the phantom into
several (for example three) sub-phantoms. If the shape of the part of the
phantom that consist not of Air (for example the patient) is rather
convex, then it should be possible to reduce the size of the outer
sub-phantoms without removing the non Air parts. In this way this
composite phantom should fit into the scanner.

Niklas
-- 
Niklas REHFELD, PhD
Projet "Quantification en Imagerie Moléculaire"
Imagerie et Modélisation en Neurobiologie et Cancérologie
IMNC - UMR 8165 CNRS
Campus d'Orsay, Bâtiment 104
15 rue Georges Clemenceau
91406 ORSAY Cedex
France

Tel : ++33 (0)1 69 15 36 46
Fax : ++33 (0)1 69 15 71 96
e-mail : rehfeld at imnc.in2p3.fr