[Gate-users] GATE SPECT simulation: less counts per projection then expected

[Niels van der Werf]

Dear all,

For my current project I’m trying to simulate the acquisition of the Jaszczak phantom filled with 100 MBq Lu-177 on a Symbia SPECT system. For this, I use the VIP-network (http://vip.creatis.insa-lyon.fr <http://vip.creatis.insa-lyon.fr/>) where the amount of primaries are calculated as time per projection (30s) times initial activity. This results in 30x100 000 000 = 3 000 000 000 primaries. 

For the Lu-177 activity, I use the MIRD radionuclide data and decay schemes, 2nd edition, KF Eckerman & A Endo Eds, The Society of Nuclear Medicine, Reston, VA, SNM, 2008, as proposed in several GATE examples. 

When this experiment is performed in real life, we get about 14000 counts per slice in 30 seconds for the highest photo peak . However, the simulation only gives about 1400 counts per slice. So, only about 10% of the counts are given for the experiment. 

Does anyone have a suggestion where this difference can come from? Thanks in advance for your time!

Best regards,
Niels van der Werf 
Erasmus MC - The Netherlands
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opengatecollaboration.org/pipermail/gate-users/attachments/20180426/64aaa258/attachment.html>