[gate-users] Energy Spectrum Woes - Dual Head and 4 Head Designs

Christian MOREL christian.morel at epfl.ch
Thu Sep 2 15:12:54 CEST 2004

Dear Ange,

I do not know exactly what detector configuration you are simulating,  
but at least from the point of view of physics, both spectra you have  
displayed in your message make some sense to me.

In the photon One spectrum, you get three peaks at about 180 keV, 511  
keV and 690 keV. Obviously, the largest peak corresponds to the sum of  
the two lowest peaks: for some reasons you detect simultaneously two  
photons of 180 keV and 511 keV in your device. Now, the lowest peak   
around 180 keV might very well correspond to Compton backscatter  
(annihilation photons scattered with 180 degrees have about 170 keV).  
Due to the scattering cross-section, back scattered photons can build  
up a peak nearly above the backscattered photon energy. This happens  
e.g. if you get a lot of material around your detector.

In the photon Two spectrum, you simply do not observe this backscatter.  
Therefore I imagine you have a different configuration for your  
detectors One and Two, in particular with regards to surrounding  

Hope this helps a bit ;-)

Best regards,

Christian Morel

Le jeudi, 2 sep 2004, à 04:40 Europe/Zurich, Angela M K Foudray a écrit  

> Recently my labmates and I have been looking into two head and 4 head
> PET system designs using LSO crystals for scintillators, F18 in water
> phantoms for the source+phantom... I can give you more details if you
> like... but we have been increasingly alarmed over the energy spectra
> that we have been getting.  We have been looking at the energies output
> in the Coincidence file of the two photons seperately.  We have been
> getting very confounding results.  First off, the two spectra are not
> "the same" (and of course, I don't mean small differences you're going
> to get from Monte Carlo) - they're just not similar.  There is a high
> energy peak near 700 keV and a low energy peak near 200 in one of them
> that we can't explain, and there's a compton peak near 350 keV in only
> the one that doesn't have the strange 200 keV peak.  If there was a
> weird 700 keV peak and a weird 200 keV peak in *both* of them, I would
> be slightly less worried - maybe - but the spectra just are completely
> different.  I understand that "photon one" (the first one listed in the
> coincidences file) is simply the first photon of a pair that, with a
> particular time resolution, happened to be "detected" first - with a
> great number of counts, the spectrum from photon one should be, to the
> eye, nearly exactly the same as the energy spectrum for photon two,
> statistically.
> I have attached a picture of the two energy spectra for a particular
> simulation (they *always* come out the same, the spectrum for photon  
> one
> has a weird low energy peak and small compton peak, as well as a weird
> high energy peak) displaying the current bane of our simulations.
> What is going on?
> If you have *ANY* insight... hit me with it!
> Ange
> <problems_energy_spectrum2.TIF>________________________________________ 
> _______
> gate-users mailing list
> gate-users at lphe1pet1.epfl.ch
> http://lphe1pet1.epfl.ch/mailman/listinfo/gate-users

Christian MOREL, PD, PhD
CH-1015 Lausanne

phone: +41 21 693 04 76
fax: +41 21 693 04 77
mobile: +41 79 292 80 70
email: christian.morel at epfl.ch
web: http://www-lphe.epfl.ch/~PET/

More information about the Gate-users mailing list