[Gate-users] History-by-history error estimation in the fixed forced detection actor

[Simon Rit]

Hi,
I have implemented this. I have not changed the method because there is a
deterministic part. I see the problem as is: you have a measure which
varies stochastically in each pixel (because there is a stochastic part in
the computation) and I estimate the uncertainy using the history-by-history
method.
So N is the number of particles during the run (mNumberOfEventsInRun) and X
is the measure (Compton scatter, Rayleigh scatter, total scatter etc.) in
one pixel. This is computed in this function
<https://github.com/OpenGATE/Gate/blob/develop/source/digits_hits/include/GateFixedForcedDetectionFunctors.hh#L1016>.
Note that all the squared and sums are computed with ITK filters in the
code, look for mIsSecondarySquaredImageEnabled in
GateFixedForcedDetectionActor.cc
<https://github.com/OpenGATE/Gate/blob/develop/source/digits_hits/src/GateFixedForcedDetectionActor.cc>.
The code is not very easy to read but that's just mimicking the algorithm
in the reference you sent for complete projections (all pixels).
I would suggest to use the batch approach to validate it if you have any
doubt.
Simon

On Sat, Jun 24, 2017 at 11:59 AM, Triltsch, Nicolas <nicolas.triltsch at tum.de
> wrote:

> Dear Gate fellows,
>
> My question aims at the history-by-history method which is used to
> calculate statistical uncertainties for the scatter images (Compton and
> Rayleigh) in the fixed forced detection actor. I have troubles to
> understand how the method  is applied within this algorithm which combines
> statistical calculation of scatter points within the object using classical
> Monte Carlo (MC) simulations (stage 1) with deterministic calculations of
> the probability to reach a certain detector pixel (stage 2).
>
> What are the parameters N and X (see the original paper I have appended)
> in our case which are required to calculate the variance for the new method?
>
> Uncertainty should only play a role in the first stage of the simulation
> where classical MC is used, however, in the second stage purely
> deterministic calculations are used. How is this considered within the
> uncertainty method?
>
> I hope that someone can help me to understand this in more detail!
>
> Thanks
>
> Nicolas Triltsch
>
> --
> B.Sc. Nicolas Triltsch
> Masterand
>
> Technische Universität München
> Physik-Department
> Lehrstuhl für Biomedizinische Physik E17
>
> James-Franck-Straße 1
> 85748 Garching b. München
>
> Tel: +49 89 289 12591
>
> nicolas.triltsch at tum.de
> www.e17.ph.tum.de
>
>
> _______________________________________________
> Gate-users mailing list
> Gate-users at lists.opengatecollaboration.org
> http://lists.opengatecollaboration.org/mailman/listinfo/gate-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opengatecollaboration.org/pipermail/gate-users/attachments/20170626/54dbd384/attachment.html>