[Gate-users] Towards Gate 9.0 (Report of GATEv8.2 bugs)

Mon Mar 2 08:01:59 CET 2020

Hi Zhengzhi,

On Thu, Feb 27, 2020 at 9:01 PM Zhengzhi Liu <zliu36 at stanford.edu> wrote:

> Dear David,
>
> Can I share some problems I have with Gate 8, which may be caused by
> bugs?  I believe I am not the only GATE user who came across these issues.
>

sure ;) ! (If bug is confirmed, an issue should be filled on github)

>
> Problem 1: * Coincident sorter*
> On top of the imcompatibility with genericRepeater as reported by Han Gyu,
> the built in coincident sorter is problematic in realistic situations. It
> works fine under ideal conditions as described in the Gate user manual
> v8.0. Taking the *cylindricalPET* as example, a cylindricalPET system can
> have 5 hierarchies at most, i.e., sector, module, submode, crystal, layer.
> This is to build an simplistic PET system. However, in real life, the
> crystal may have a wrapper, and detectors may be placed in some supporting
> structure and shielding. When I insert any other geometry into the system
> (for example, I added an aluminum wrapper for crystals, whether I attach it
> the system or not ), the coincidence sorter would fail immediately, but
> singles generation works fine.
>

a first workaround has been proposed recently by Christian Pommaranz, see
https://github.com/OpenGATE/Gate/pull/297 and
https://github.com/OpenGATE/Gate/issues/296. Could you try if it helps your
issues? Please, note that the coincidence sorter does not work for
combinations of angularRepeater and genericRepeater for now.

Please let us know if it helps ...

> Problem 2: *work with voxelized source and Phantom*
> First, I can't get the built in example code (benchImageNavigators/pet)
> work. Second, it looks like no body has ever succeeded in visualing their
> voxelized source. The earliest email
> <http://lists.opengatecollaboration.org/pipermail/gate-users/2010-April/004126.html> about
> voxelized source visualization can date back to *10 years* agao, could
> you make an official reply to that or update it in the upcoming version?
>

Please, could you please describe a bit more the first error?

Second, for the visualisation of the source, as proposed by Christopher and
Matthew you may add an image at the exact same place.

> Thank you very much for your time reading and considering the problems.
>

thanks to you for reporting !

David
PS: I will be travelling this week, back on track next week ...

> Sincerely
>
> Zhengzhi
>
>
>
> On Wed, Feb 19, 2020 at 7:21 AM David Sarrut <
> David.Sarrut at creatis.insa-lyon.fr> wrote:
>
>>
>> excellent, thanks to both of you!
>>
>> On Wed, Feb 19, 2020 at 12:23 PM 강한규 <lovehangulp at naver.com> wrote:
>>
>>> Dear David, and Christian,
>>>
>>>  Thanks for your kind response.
>>>
>>>  For the bug#1: I uploaded an issue about the GATE bug with
>>> cubicArray.(issue#296)
>>>
>>>  For the bug#2: Christian (included in this email) will handle the
>>> coincidence sorter issue wit his own pull requests (PR) on the git hub.
>>>
>>>  Best regards,
>>>
>>>  Han Gyu
>>>
>>> --------
>>>
>>> Han Gyu Kang, Ph.D, Researcher
>>>
>>> National Institute of Radiological Sciences (NIRS)
>>> National Institutes for Quantum and Radiological Science and Technology
>>> (QST)
>>>  4-9-1 Anagawa, Inage-ku, Chiba 263-8555, JAPAN
>>>
>>>
>>>
>>> -----Original Message-----
>>> *From:* "David Sarrut"<David.Sarrut at creatis.insa-lyon.fr>
>>> *To:* "강한규"<lovehangulp at naver.com>;
>>> *Cc:* "gate-users"<gate-users at lists.opengatecollaboration.org>;
>>> *Sent:* 2020-02-19 (수) 17:25:22 (GMT+09:00)
>>> *Subject:* Re: [Gate-users] Towards Gate 9.0 (Report of GATEv8.2 bugs)
>>>
>>> Hello 강한규,
>>>
>>> thanks very much for the very clear report!
>>>
>>> For bug#1: I am not expert in cubicArray repeater (I wrote the
>>> genericRepeater long time ago). We should investigate, but not sure it will
>>> be done for 9.0. Can you please put your email as an issue on github so
>>> that we keep memory of it?
>>>
>>> For bug#2: sorry, I cannot remember of the patch you mention (I probably
>>> forget this, my bad). Could you please provide the patch as a PR ? It will
>>> be included rapidly!
>>>
>>> Thanks a lot, do not hesitate to ask if need help for the PR
>>>
>>> Thank again,
>>> David
>>>
>>>
>>> On Wed, Feb 19, 2020 at 9:08 AM 강한규 <lovehangulp at naver.com> wrote:
>>>
>>> Dear David Sarrut,
>>>
>>>
>>>
>>> Let me share the bugs of GATEv8.2.
>>>
>>> I tried to simulate a *PEM (Positron emission mammography)* scanner
>>> with GATEv8.2.(slide#3-8)
>>>
>>> For the system type, "*PETscanne*r" was used because of the non-fixed
>>> geometry.
>>>
>>>
>>>
>>> *Bug#1: Check pattern with the cubicArray repeater.*
>>>
>>> When the back-to-back source (red) is at the center of the scanner,
>>> there was no problem.
>>>
>>> However, as the source was placed close to the one detector side, the
>>> check pattern was appeared on the detector response.(slide#10)
>>>
>>> This problem can be solved by using genericRepeater.(slide#11-12)
>>>
>>> Unfortunately, however, the coincidence sorter does not work with the
>>> genericRepeater.
>>>
>>>
>>>
>>>
>>>
>>> *Bug#2: Coincidence sorter does not work with the genericRepeater
>>> (slide#13-16)*
>>>
>>>          The coincidence sorter issue could be solved by using git patch
>>> which was proposed by Christian Pommaranz (
>>> pommranz at astro.uni-tuebingen.de).
>>>
>>>          After several steps of GATE macro and source code modification,
>>> the PEM scanner simulation worked successfully.(slide#16)
>>>
>>>          However, for the beginners of GATE, it is going to hard to do
>>> it by themselves.
>>>  Hopefully, the abovementioned bugs would be resolved in the GATEv9.0.
>>>
>>>  Best regards,
>>>
>>>
>>>
>>> Han Gyu Kang, Ph.D.
>>>
>>>
>>>
>>> Imaging Physics Group, NIRS-QST, Japan.
>>>
>>>
>>>
>>> -----Original Message-----
>>> *From:* "David Sarrut"<David.Sarrut at creatis.insa-lyon.fr>
>>> *To:* "gate-users"<gate-users at lists.opengatecollaboration.org>;
>>> *Cc:*
>>> *Sent:* 2020-02-19 (수) 16:28:07 (GMT+09:00)
>>> *Subject:* [Gate-users] Towards Gate 9.0
>>>
>>> Dear all,
>>>
>>> We are currently preparing the next Gate release, version 9.0, based on
>>> the current Geant4 release 10.6. This version includes a bunch of novelties
>>> compared to the previous 8.2 version (hence the jump in version number).
>>>
>>> Please, if you want to include new feature or bug correction in this
>>> version, *submit your PullRequest (PR) asap*. We usually ask to have
>>> one PR with source code only in the main Gate repository, another PR with
>>> one example in the GateContrib repository, and some documentation changes
>>> via readthedocs <https://opengate.readthedocs.io/en/latest/>. In any
>>> case, do not hesitate to contact us for any help, we are happy to help !
>>>
>>> *For all*: please we need help to perform tests. Gate is a large beast
>>> and we still do not have a proper unit test system (working on it). This is
>>> simple: download and compile the branch "develop" and test it with your own
>>> simulation to see if anything is fine. Please, report us is anything seems
>>> strange. On our side, we have some benchmarks but they do not cover all the
>>> parts of Gate. Continuous integration
>>> <https://en.wikipedia.org/wiki/Continuous_integration> on github via
>>> Azure is currently being implemented (thanks to Thomas!) and will be ready
>>> soon.
>>>
>>> In advance, many many thanks to all recent contributors: Fatima, Anders,
>>> Mathieu, Thomas, David, Marta, Loic, Sam, Christian, Mateuz etc ... I am
>>> sure I forget people, sorry ... I will properly update the list of
>>> contributors in the following weeks :)
>>>
>>> The ETA for the next release is early March.
>>>
>>> David
>>>
>>>
>>>

-- 
David Sarrut, Phd
Directeur de recherche CNRS
CREATIS, UMR CNRS 5220, Inserm U1206
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/pipermail/gate-users/attachments/20200302/fe490a25/attachment-0001.html>