[Gate-users] Issues on voxelized dose calculations for a phantom (using Hounsfield converter)

Mon Jun 1 11:07:06 CEST 2015

Hello Alex,

pull-request #23 is now merged, thanks again. I will ask for a
configuration in the dashboard <http://my.cdash.org/index.php?project=GATE>
to be compiled daily with cxx11 enabled.

David

On Thu, May 28, 2015 at 3:14 PM, Alex Vergara Gil <alexvergaragil at gmail.com>
wrote:

> Dear David
>
> On pull request https://github.com/OpenGATE/Gate/pull/23 I added both
> the changes to make Gate c++11 compatible and the fix of the mentioned
> bug.
>
> Regards
> Alex
>
> 2015-05-28 8:13 GMT-04:00, Alex Vergara Gil <alexvergaragil at gmail.com>:
> > Dear David
> >
> > There is another things, code 'elegancy' lead to some specific
> > enhances that must be taken in consideration:
> >
> > 1. Indentation: helps to read better the code so bugs and unnecessary
> > sentences raise from all their hidden places, specially when you make
> > automatic indentation.
> > 2. changing from i++ to ++i as I explain leads to optimization of code
> > since it avoid a whole extra step at each call.
> > 3. Using std::vector functionalities lead to code enhancement since it
> > use a really optimized method for travel among lists.
> > 4. Using structured vectors is a security enhancement since you can no
> > longer create vector from different sizes, avoiding to not initialize
> > some of them and the corresponding segmentation fault when calling.
> >
> > I know there are a bunch of changes, but I can run all my macros with
> > them, decreasing a little the memory consumption, increasing a little
> > the histories per second and finally when I look now at the code it
> > seems made in modern times ;)
> >
> > Regards
> > Alex
> >
> > 2015-05-28 7:24 GMT-04:00, Alex Vergara Gil <alexvergaragil at gmail.com>:
> >> Hello David
> >>
> >> As I said on github, you can create a testing branch so I can still
> >> contribute with great amount of changes and user can test them and
> >> report bugs, this will not affect Gate at all and we all be happy ;)
> >>
> >> Regards
> >> Alex
> >>
> >> 2015-05-28 2:11 GMT-04:00, David Sarrut
> >> <David.Sarrut at creatis.insa-lyon.fr>:
> >>> Hello Alex,
> >>>
> >>> as said on github (see https://github.com/OpenGATE/Gate/issues/22),
> >>> could
> >>> you please provide a separate pull-request for this issue, *without*
> >>> replacement of std::vector -> please keep it simple. Even if the code
> is
> >>> not so good, I acknowledge that ;) refrain to changing it to be more
> >>> 'elegant' as it could lead to unexpected issues.
> >>>
> >>> dont get me wrong, I really appreciate your contribution, but we need
> to
> >>> ensure to not break something ...
> >>>
> >>> thanks,
> >>> David
> >>>
> >>> On Tue, May 26, 2015 at 2:02 PM, Alex Vergara Gil
> >>> <alexvergaragil at gmail.com>
> >>> wrote:
> >>>
> >>>> Dear David
> >>>>
> >>>> Indeed, doing that replace solved the bug, not only for range
> >>>> translator but also for hounsfield translator, in my pull request are
> >>>> the necessary changes with a bunch of code enhacements (basically a
> >>>> lot of replacements to use std::vector functionalities).
> >>>>
> >>>> Look now at one of my results on a phantom compared with FLUKA.
> >>>>
> >>>> Regards
> >>>> Alex
> >>>>
> >>>> 2015-05-26 3:32 GMT-04:00, David Sarrut
> >>>> <David.Sarrut at creatis.insa-lyon.fr
> >>>> >:
> >>>> > Hello Alex,
> >>>> >
> >>>> > I looked at your coments on the range translator and indeed, there
> is
> >>>> > a
> >>>> > bug. Could you please test if replacing the line
> >>>> > if(r2> pImage->GetOutsideValue()+1){
> >>>> > with
> >>>> > if(r2>= pImage->GetOutsideValue()+1){
> >>>> >
> >>>> > solve the issue ?
> >>>> >
> >>>> > Thanks you indeed to propose me to be coauthor of the paper, but I
> >>>> > really
> >>>> > did nothing to this work, so I cannot accept, this is not fair ;)
> >>>> >
> >>>> > I am however interested to read the draft if possible.
> >>>> >
> >>>> > As journal, I can suggest maybe : http://jnm.snmjournals.org or
> >>>> > http://www.springer.com/medicine/nuclear+medicine/journal/259
> >>>> >
> >>>> > David
> >>>> >
> >>>> >
> >>>> >
> >>>> > On Thu, May 21, 2015 at 1:34 PM, Alex Vergara Gil
> >>>> > <alexvergaragil at gmail.com>
> >>>> > wrote:
> >>>> >
> >>>> >> Dear David
> >>>> >>
> >>>> >> I think I have solved the bug, I ran several tests and I think
> >>>> >> results
> >>>> >> are fine now for both range and hounsfield translators.
> >>>> >>
> >>>> >> Next days I will be writing my results. If you want I will send you
> >>>> >> the draft I will prepare so you can coauthor the paper, is there a
> >>>> >> magazine you may recommended?
> >>>> >>
> >>>> >> Suggested title: "Using Gate for Molecular Radiotherapy Dosimetry
> on
> >>>> >> clinical environments"
> >>>> >>
> >>>> >> Regards
> >>>> >> Alex
> >>>> >>
> >>>> >> 2015-05-20 1:12 GMT-04:00, David Sarrut
> >>>> >> <David.Sarrut at creatis.insa-lyon.fr
> >>>> >> >:
> >>>> >> > Hello Alex,
> >>>> >> >
> >>>> >> > I will study this issue probably next week,
> >>>> >> >
> >>>> >> > thanks again for reporting,
> >>>> >> > David
> >>>> >> >
> >>>> >> > On Tue, May 19, 2015 at 10:49 PM, Alex Vergara Gil
> >>>> >> > <alexvergaragil at gmail.com
> >>>> >> >> wrote:
> >>>> >> >
> >>>> >> >> Dear David
> >>>> >> >>
> >>>> >> >> I have modified my macros and now I have a good patient dose
> >>>> >> >> image,
> >>>> so
> >>>> >> >> I think Hounsfield Translator is free of bugs, I will try to put
> >>>> >> >> Air
> >>>> >> >> as first segmentation material to see whats happens, but in the
> >>>> >> >> way
> >>>> >> >> I
> >>>> >> >> am seeing this Hounsfield is far better approach for the
> problem.
> >>>> >> >>
> >>>> >> >> I will update my templates to achieve this macro and when is
> >>>> >> >> ready
> >>>> >> >> I
> >>>> >> >> will publish my results.
> >>>> >> >> On issue #18 I put all my progress, This image was obtained with
> >>>> >> >> geant4 10.1.1 + gate with all modifications of pull request #21
> >>>> >> >> applied.
> >>>> >> >>
> >>>> >> >> Regards
> >>>> >> >> Alex
> >>>> >> >>
> >>>> >> >> 2015-05-19 14:29 GMT-04:00, Alex Vergara Gil <
> >>>> alexvergaragil at gmail.com
> >>>> >> >:
> >>>> >> >> > Dear David
> >>>> >> >> >
> >>>> >> >> > I am trying to move my workflow to Hounsfield converter, since
> >>>> >> >> > I
> >>>> can
> >>>> >> >> > get the range translator to work properly (In Patient Phantom
> 2
> >>>> >> >> > you
> >>>> >> >> > may see that a material - Muscle - is taken as air). Up to now
> >>>> >> >> > I
> >>>> >> >> > have
> >>>> >> >> > successfully ran a voxelized physical phantom as you may see
> in
> >>>> >> >> > the
> >>>> >> >> > image (cylinder phantom 1), but when using a patient phantom
> >>>> >> >> > thing
> >>>> >> >> > goes really odd! Somehow geometry are no longer rendered
> >>>> >> >> > correctly
> >>>> >> >> > (patient phantom 1). I am trying to check what is wrong with
> >>>> >> >> > the
> >>>> >> >> > images but I see no errors, even the density map I export from
> >>>> >> >> > gate
> >>>> >> >> > is
> >>>> >> >> > fine (patient density phantom 1). I also see no relative
> >>>> differences
> >>>> >> >> > between the cylinder and patient macros, except for the image
> >>>> >> >> > properties.
> >>>> >> >> >
> >>>> >> >> > Regards
> >>>> >> >> > Alex
> >>>> >> >> >
> >>>> >> >>
> >>>> >> >
> >>>> >> >
> >>>> >> >
> >>>> >> > --
> >>>> >> > David Sarrut, Phd
> >>>> >> > Directeur de recherche CNRS
> >>>> >> > CREATIS, UMR CNRS 5220, Inserm U 1044
> >>>> >> > Centre de lutte contre le cancer Léon Bérard
> >>>> >> > 28 rue Laënnec, 69373 Lyon cedex 08
> >>>> >> > Tel : 04 78 78 51 51 / 06 74 72 05 42
> >>>> >> > http://www.creatis.insa-lyon.fr/~dsarrut
> >>>> >> > _________________________________
> >>>> >> >  "2 + 2 = 5,  for extremely large values of 2"
> >>>> >> > _________________________________
> >>>> >> >
> >>>> >>
> >>>> >
> >>>> >
> >>>> >
> >>>> > --
> >>>> > David Sarrut, Phd
> >>>> > Directeur de recherche CNRS
> >>>> > CREATIS, UMR CNRS 5220, Inserm U 1044
> >>>> > Centre de lutte contre le cancer Léon Bérard
> >>>> > 28 rue Laënnec, 69373 Lyon cedex 08
> >>>> > Tel : 04 78 78 51 51 / 06 74 72 05 42
> >>>> > http://www.creatis.insa-lyon.fr/~dsarrut
> >>>> > _________________________________
> >>>> >  "2 + 2 = 5,  for extremely large values of 2"
> >>>> > _________________________________
> >>>> >
> >>>>
> >>>
> >>>
> >>>
> >>> --
> >>> David Sarrut, Phd
> >>> Directeur de recherche CNRS
> >>> CREATIS, UMR CNRS 5220, Inserm U 1044
> >>> Centre de lutte contre le cancer Léon Bérard
> >>> 28 rue Laënnec, 69373 Lyon cedex 08
> >>> Tel : 04 78 78 51 51 / 06 74 72 05 42
> >>> http://www.creatis.insa-lyon.fr/~dsarrut
> >>> _________________________________
> >>>  "2 + 2 = 5,  for extremely large values of 2"
> >>> _________________________________
> >>>
> >>
> >
>

-- 
David Sarrut, Phd
Directeur de recherche CNRS
CREATIS, UMR CNRS 5220, Inserm U 1044
Centre de lutte contre le cancer Léon Bérard
28 rue Laënnec, 69373 Lyon cedex 08
Tel : 04 78 78 51 51 / 06 74 72 05 42
http://www.creatis.insa-lyon.fr/~dsarrut
_________________________________
 "2 + 2 = 5,  for extremely large values of 2"
_________________________________
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opengatecollaboration.org/mailman/private/gate-users/attachments/20150601/bf32d676/attachment.html>