[Gate-users] Problem with the coincidence

[Jinxia Guo]

Hi Simon,

Thanks for your answer. I have do some other simulation too. In all
simulation, the time resolution is set to 0 and I get the distribution of
time difference of all counts. I can't find the anything to help me
understand that the time difference of singles in a coincidences is larger
than coincidence window. According to the fig.8.5 in GATE user guide, it can
be considered as a coincidence only if the singles are in the coincidence
time window. So the time difference of singles in a coincidence shouldn't be
larger than the coincidence window. I can't understand that you said “ if
you have a sort a gaussian function, or a dirac function at a certain time
difference value”. Can you explain for me?

For the " multiple coincidences" problem, there are 9 policies to choose and
I choose the default policy in the simulation. I can understand the result
now. But I have another question. The policy "take Winner Of Goods" means
"Only the good sub-pair with the highest energy is taken". Why choose the
highest energy subpair? To avoid scatters?

Thanks,
Jinxia


在 2010年4月19日 下午5:09，Simon Stute <gate.stute at gmail.com>写道：

> Hi,
>
> 2010/4/19 Jinxia Guo <jinxia.guo2009 at gmail.com>
>
> Dear gate users,
>>
>> I encountered a problem when I inspection the time difference of two
>> singles
>> in a coincidence. In the simulation, there are many coincidences in which
>> the time difference of two singles is larger than the coincidence window.
>> I
>> do two simulation. In the first one, the time resolution is set to 200ps
>> and
>> the coincidence window is set to 200ps. In the second one, the time
>> resolution is set to 0 and the coincidence window is set to 200ps. In both
>> simulation, the problem exists. I can't find why this happened with
>> chapter8
>> of GATE user guide. So who know how gate does its coincidence？Why the time
>> difference of singles in coincidences is larger than the coincidence
>> window?
>>
>>
>
> You can plot the time difference, in order to have an idea of what the
> problem really looks like, it would be a good start.
> For example to see if you have a sort a gaussian function, or a dirac
> function at a certain time difference value, it would be easier to try to
> understand the problem.
> And to study the problem, just set the time resolution to 0, otherwise it
> would be too difficult to analyse the results.
>
>
>
>> In addition to the above simulation, I do another two simulation. In first
>> one, the time resolution is 300ps and coincidence window is 10ns. In
>> second
>> one, the time resolution is 300ps too and coincidence window is 100ns. I
>> found there are more coincidences in the first simulation than in the
>> second
>> simulation,about five times larger.Then who can explain it？
>>
>>
>
> By setting your time window to 100ns, you will have a lot of multiple
> coincidences, so it depends on what policy you choose.
> By default, takeTheWinner is chosen, so it explains why you get five times
> less coincidences in your second simulation.
> See the user's guide.
>
> Cheers,
> Simon
>
>
>> By the way, in all simulations, the diameter of simulated PET ring is
>> about
>> 82cm. And the simulation time is the same.
>>
>> Hope for help. Thank you!
>>
>> -
>> Jinxia Guo
>> School of Life Sci&Tech
>> D11 Building Room 503
>>
>> Huazhong Univ. of Sci. & Tech
>>
>> 1037 Luoyu Road, Wuhan 430074, China
>> Tel:+86-27-8779-2294
>> _______________________________________________
>> Gate-users mailing list
>> Gate-users at lists.opengatecollaboration.org
>> http://lists.opengatecollaboration.org/mailman/listinfo/gate-users
>
>
>


-- 
Jinxia Guo
School of Life Sci&Tech
D11 Building Room 503

Huazhong Univ. of Sci. & Tech

1037 Luoyu Road, Wuhan 430074, China
Tel:+86-27-8779-2294