Re: [fluka-discuss]: Conversion of ICRP No 110 Phantoms to .vxl?

From: Vittorio Boccone <dr.vittorio.boccone_at_ieee.org>
Date: Thu, 7 Jul 2016 20:30:51 +0200

Dear Alex,
always keep the fluke-list in cc as more people might benefit of this discussion

> On 07 Jul 2016, at 09:24, Ševčik Aleksandras <aleksandras.sevcik_at_ktu.edu> wrote:
>
> Thank you, Vittorio. I'm a master student in Medical Physics; what we are taught it's doesn't cover any of these topics neither programming. We are seen more like users only while I don't like this and want to learn mc transport coding myself and get out of the black box effect, if I may say so. However, this means that I have poor or no skills in some of the areas which result in asking basic/obvious questions as manual or tutorials are not always clear for me. So any respond for experts are highly valuable to me. In the whole country we have only several dozens of medical physicists and not a single of them have experience in mc transport coding. In general, Lithuania only seeks to become an associated CERN member so there are so many areas to improve here..
You are right, university doesn’t teach much about the computing tools and programming. But every student in the STEM fields must get the hands on a scripting/analysis language which can help him/her in computational or analysis tasks and in general to solve more or less complex tasks

> So far I understood the schematics. I have .dat file, with the help of writeVoxelFile.f I can convert it to .vxl as per your instructions ( I found it in the manual finally too). I also understand that I can open writeVoxelFile.f and edit PARAMETERS (DZ...NZ) if required. So now we have voxel geometry where similar voxels account for the same region. Let's say we get 3000 such regions.
If you are beginner. I would suggest you to go thought the manual introduction and the slides of the beginner course at first. A voxel is a regular 3D pixel structure with NX, NY, NZ pixels extension over the three dimensions. in the dat file you find a table of organ indexes which must be converted in a voxel structure. This is what the writeVoxelFile does. The dat file contain the ascii table with the organ index per each pixel.

You need to pass from here in order to make the voxel file. All you need is in the repo, you just have to adapt the NX,NY,NZ, DX, DY, DZ variables to the values that mentioned in the Readme file, compile and run the conversion. 3000 different tissue regions are an overkill. Many people reduces them to a few hundreds at most.

> So now we need to assign mat to these regions. I assume that definitions of the materials are encoded in the rest of .dat files and the scripting language like Python can be used. Unfortunately I have no any use or experience of any programming language before.
>
> The following questions arise:
>
> 1) You mentioned I could it manually. Would you be so kind to explain this? Or should I better go and learn Python and come back after 2-3 weeks when I learn the basics? I planned to do this sooner or earlier anyway.
IMO a python course is a good idea only if it contains hands-on exercises.

> 2) Interesting : so in .vxl file we have voxels that are differentiated into the regions according levels of grey (HU units)? So we appoint materials(compounds) according HU values?
The voxel file contains the region/organ index for each pixel one after the other. What the organ/region is should be added in the input file with the user cards. You can use the flair interface to build and check your file.

> One more 3) question. As I understand Flair has automatized all these actions when import DICOM and convert it to .vxl. , while using materials.inp where 47 compounds are described. However there is no automatic adding of corresponding material => voxel (region). In other words, I have predefined materials imported ("HU<1020", HU<"1000", etc.) and at least 3000 voxel regions (VOXEL001...VOXEL3000). But how should I choose for which region which "HU<1020" should be selected? I read some topics in the mailing list about this issue but no clear answer was given.
Did you already looked at this thread (and references therein)?
http://www.fluka.org/web_archive/earchive/new-fluka-discuss/8841.html



Best,
Vittorio

__________________________________________________________________________
You can manage unsubscription from this mailing list at https://www.fluka.org/fluka.php?id=acc_info
Received on Thu Jul 07 2016 - 22:14:45 CEST

This archive was generated by hypermail 2.3.0 : Thu Jul 07 2016 - 22:14:52 CEST