[gate-users] Counts vs. time slice

[Claude Comtat]

Hi Corinne,

Corinne J Groiselle wrote:

>
>
> Angela M K Foudray wrote:
>
>>That's good to know, but unfortunately, I'm looking to specify the
>>number of Coincidence events - the number of annihilations should be the
>>same as the time multiplied by the source strengths added up, right? 
>>  
>>
> Not exactly... Because of the way the particle generator works the 
> number of annihilations is not exactly the time by the activity. So 
> it's a way to be sure of the number of emitted particles.


What do you mean by "not exactly" ? The Poisson nature of the decay, or 
any systematic bias ? In the PET benchmark, you have a comparison with 
the expected number of annihilations :

Activity*(StopTime - StartTime)*DecayFactor

where DecayFactor accounts for the decay up to StartTime and the decay 
during the acquisition. This figure is in pretty good agreement with the 
actual number of generated events, the difference coming from the 
Poisson nature of the decay, explicitly simulated. In that sense, it is 
"not exactly" the time by the activity.

Regards,

Claude