[gate-users] Question: Using GATE for non cylindrical PET systems, with non scintillation detectors

Miguel Couceiro couceiro at lipc.fis.uc.pt
Thu Jan 12 01:49:11 CET 2006

Hi Sebastien,

Thank you for your answer.

One of the things we want to study is the more efficient scanner 
geometry for our detectors. For this, we want to test three basic 
geometries (parallelepipedic, pentagonal and hexagonal) and some 
variants of the pentagonal and hexagonal geometries, with some degree of 
detector overlap. We also may consider other geometries, but for now  
these ones will be the starting point. The scanner will have a 240 cm 
FOV axial length, and 80 cm inner cylindrical radius.

The detectors are based on RPCs gasos detectors. These detectors have a 
very good spacial resolution (depth of interaction included), very fast 
response (much less than crystal based detectors), very good time 
resolution for TOF information, and very small dead time.

Due to the construction of the RPC detectors, each detector must be made 
parallelepipedic. Their dimensions will depend on several factors, as 
electronic complexity, total cost of the scanner, easy of assembly, etc.

In spite of the very good spatial resolution of each detector, we want 
to limit it to a mesh of 1 or 2 mm in the plane parallel to the scanner 
axis. So, we must parametrize the scanner in order to have a detection 
grid with 1 or 2 mm.

As I mentioned, we don't want to simulate the processes in the 
detectors, just parametrize them in terms of detection efficiency, time 
resolution and dead time.


Sebastien JAN wrote:

> Hi Miguel,
> Could you explain how is your detector geometry please ?
> Sebastien
> Miguel Couceiro wrote:
>> Hi all,
>> We are performing some simulations, using GEANT4, to test the 
>> performance
>> of an all body PET system using a new kind of detectors.
>> The problem is that, with these detectors, the geometry can not be
>> cylindrical. Besides, these detectors are not based on scintillation and
>> do not have energy resolution (but their efficiency with energy is such
>> that they work as if they have a lower energy threshold of 300 keV).
>> Since we already know the detector characteristics we are just 
>> interested
>> in parameterizing the detectors in order to know if an hit takes place,
>> and where.
>> Performing the simulations in GEANT4 is tedious, because we have to code
>> all the time related tasks, as well as the coincidence events and
>> sinograms.
>> I was wandering if it is possible to use GATE to simulate a PET 
>> system of
>> this kind, taking advantage of the GATE capabilities to deal with 
>> time, as
>> well as the output files.
>> Thanks,
>> Miguel
>> _______________________________________________
>> gate-users mailing list
>> gate-users at lphe1pet1.epfl.ch
>> http://lphe1pet1.epfl.ch/mailman/listinfo/gate-users
> _______________________________________________
> gate-users mailing list
> gate-users at lphe1pet1.epfl.ch
> http://lphe1pet1.epfl.ch/mailman/listinfo/gate-users

More information about the Gate-users mailing list