[Gate-users] voxellized phantom attenuation problem

[ayanas at mail.med.upenn.edu]

Dear Gate'rs

I am observing a problem with a voxelized phantom (using compressed voxel)
with gate-3.1.1.
Here is what happens:
I load a voxellized MCAT phantom as the attenuating medium.
The source is a simple point source at the center of the phantom.  

I am recording the un-scattered photons (140 keV) making out 
of the MCAT torso. 

Obervations:
1) When I make the material definitions of the all voxels as "Bone", 
(a uniform material made of bone) I get a sensible number of photons
un-scattered. This is fine..

2) When I make the material definitions of the all voxels as, "Lung"
I, again,  get a sensible number of photons un-scattered. This is also fine..

3) However, When I make the original MCAT definition, that is with
Body (muscle),  Lung, Bone etc, I get a number of 
un-scattered photons much higher than the "All Bone" case 
making out of the MCAT torso; 
while I would expect to get a number between "All Bone" and "All Lung" case, 
(the densest material is the Bone, and the least dense is Lung). 

Has anybody observed such a behaviour? 
Any ideas/suggestions as to what my problem might be? 

I also saw a patch for gate 3.1.2 (patch_v3.1.2.tar.bin) with a mention
to "compressed voxel capabilities". Does (Should) this patch apply
also to gate 3.1.1 ? 

Thank you very much,
ahmet.

-- 
Ahmet Sedat Ayan, Ph.D

Department of Radiology
University of Pennsylvania 
Philadelphia, PA 19104

Voice      : (++ 1 215) 349-5751 (W) (GMT-5)