[Gate-users] unwanted scatter events in primary spectrum

[Samuel BURG]

Hi Gaters,


I have a crazy issue with scattering and primary spectra (SPECT simulation with 
Tc99m)

when I create a primary spectrum from root data save from a gate simulation 
there are to many events out of the dedicated energy spike, and it look like 
event tagged as primary events are in fact scatter events.

this occurred in numerous different situations/configurations. I haven't find 
any configuration who doesn't produce such a problem, even with the SPECT 
benchmark.

here's all what I have tested, nothing solving the issue :
- 32 / 64 bits arch
- single CPU / SMP / Cluster
- GATE version 4.0 / 5.1
- source shape : point / cylinder
- collimator : with / without
- PMT back compartment : single volume / complex multi volumes
- disabling all  non-crystal volume (setting all material to "Air", or removing 
all volumes from simulation)

- different energyCut
- crystal material : NaI / NaI(Tl)
- with very basic digitizer (only adder) or very complex digitizer (all SPECT 
modules but optic tracking)
- same issue with the benchmark (root output and ASCII output)
- same results using different fields of the root file to asses with event is 
primary or scatter : RayleighVolName, RayleighPhantom, RayleighCrystal, 
comptVolName, comptonPhantom, comptonCrystal

Software versions :
OS : GNU Linux, Ubuntu 10.4 LTS, x86_64
GATE : 4.0 & 5.1
root : 5.20/00      24 June 2008
clhep : v2.0.3.2
geant4 : 4.9.1.p02
gcc : 4.2.x (replacing system default 4.4.x)

What's crazy is that in the sample file 
gate_v4.0.0/benchmarks/benchmarkSPECT/Results_benchmark.doc the primary spectrum 
looks like a primary  spectrum, and when I run the benchmark, the primary 
spectrum has some additional scatter event !

So, there is something I really don't understand here, and it turns me crazy !

Attached :-> too big files, not attached, use sendspace links below
- the outputs of the SPECTbenchmark compressed with 7z, ascii files and root 
files.
- screen capt. of the sprectra I have with my own processing and with the root 
processing of the benchmark (root output)
- results of the processing of ASCII output

someone has an idea before I break down ?

Best regards,

Samuel

Link for files on sendspace (ask me if you have any trouble downloading the 
files):
http://www.sendspace.com/file/vo7sa0 
http://www.sendspace.com/file/0ofue2 
http://www.sendspace.com/file/nvw4ef 
http://www.sendspace.com/file/w1m0va 
http://www.sendspace.com/file/seshz9 
http://www.sendspace.com/file/xck3rt 

-----------------------------------------------------------
Samuel Burg


Mail : <samuel_burg at yahoo.fr>
-----------------------------------------------------------
"Aw, come on, who would be so stupid as to insert a cast
to make an error go away without actually fixing the error?"
Apparently everyone. 
-- Raymond Chen  --
-----------------------------------------------------------