Hi guys,
I was running some simulations on a DICOM file related to the lungs of a patient in order to score the dose released to the matter, in this case a human body flooded in a volume of surrounding air. Now I read on this forum that DOSE-EQ (USRBIN) card is related to H^*(10) while I’m employing a photon beam of 25MeV that goes far over the 10 cm at which the H^*(10) standard should be considered. Here comes my first question:
could be acceptable to you to leverage the DOSE-EQ card in this case?
Secondly the unit of measure, I read in this forum that DOSE-EQ card produces pSv \cdot cm^3 (I have a 3D geometry) and therefore should be normalized by the volume (i.e. divided) of the model. Now the second question:
using the normalisation box and inserting what I assumed the volume of the entire geometry: 70500\thinspace cm^3 (I took this value from the DICOM interface, under the “information box” just before the voxelization procedure) the scale seems to increase a lot. I’m a bit confused could that be that in this case where I have a voxel file I don’t need to normalize the results? (for instance, putting into the norm box 1/70500\thinspace cm^{-3} totally nulls the dose, which I think is pretty unphysical!)
do I have to normalise the results by volume? So from the second screenshot, the red area shows a value of \approx 1000\thinspace pSv
The results are already normalised by the initial energy beam? Leading to the blue area of the DICOM interface to report a value of 25MeV \cdot 0.3 \rightarrow 7.5 \thinspace pSv (that looks a bit too low to my expectations)
For materials: material.inp (12.1 KB) Official flair site (the dicom.tgz file), here I got the body.mat, that I cant link directly because the *.mat unfortunately seems not to be supported by the forum engine.
First, based on the input files and pictures it seems, that you are not using the FLUKA version released by CERN at https://fluka.cern/
Please note, that any technical support will be provided only for this release on this forum.
I don’t think DOSE-EQ (Ambient dose equivalent, by default) is appropriate for this application because it calculates the dose a person’s body would receive. It is achieved with fluence-to-dose conversion factors.
Since you are interested in doses inside a human body, the more appropriate quantity would be simply to score DOSE.
Generally FLUKA automatically normalizes the results with the volume of the region/voxel. (Except a region based USRBIN scoring), so the results will be normalized to one primary particle. There is no normalization regarding the beam energy.
For an oncological treatment plan, DOSE would be the right quantity, it calculates absorbed dose in the material specified by the geometry.
The default of DOSE-EQ is indeed H*(10).
This is an operational quantity for radiological protection in which dosimeters and monitors are calibrated. This quantity estimates the effective dose E, in which dose limits are expressed, but which cannot be measured. E is defined for stochastic radiation effects. The use of E , and derived from it H*(10) , is not appropriate for an oncological target tissue, because the doses are so high that the damage is deterministic. Their use in diagnostics is acceptable if one does not calculate individual radiation risk of a patient, but uses it for comparison of different diagnostic modalities.
In your application, absorbed dose (DOSE) is clearly the quantity you should score.
Best regards, Thomas
PS: the “10” in H*(10) stands for 10 mm, the depth in tissue in which the value of the quantity is determined. It has nothing to do with the range of the particles. The use of H*(10) for photons of 25 MeV is also a matter of discussion as a location in 10 mm depth in the body is not representative at these energies, but this is a different topic …
First of all, I would like to thank you for your answer in such short time.
Initially, I thought FLAIR was considered part of the FLUKA project at CERN. Currently, I’m using the latest version downloaded from: The official FLUKA site: About FLUKA, from what you said I suppose there is another version!
However, I changed the scoring card to DOSE (USRBIN) instead of DOSE-EQ (USRBIN), unfortunately, although the human body got in contact with a 1GeV photons beam (I did a test also with 25MeV photons) in the DICOM interface, nothing in terms of Dose is displayed. I’m trying to explain it to myself, but at the moment I think it’s because generally the dicom interface doesn’t allow to normalise the results, while in the Geometry panel, if I normalise by the volume of the total voxel geometry something interesting comes out, both for the 25MeV case and for the 1GeV (a bit unrealistic) case.
In order to overcome the issue from the DICOM “RT-Viewer” I think it could be reasonable to extract the *.dat file of the simulation and normalise it manually in Python or similar and then upload the .bnn/.dat file into RT-Viewer. But at the moment I don’t know if this path is doable or the only one if I want to display that dose to the tissue of the DICOM file. Only Dose-eq scoring seems to work in synergy with DICOM
Could I ask you what do you mean for “normalized to one primary particle”, I personally don’t get which quantity fluka is referring to honestly - maybe the number of Primaries simulated?
the Flair version released by CERN and distributed at https://flair.cern allows the normalization of results in the DICOM interface. We can’t help you with FLUKA/Flair versions that CERN did not release.
The FLUKA results are generally expressed for one simulated primary particle, e.g. dose generated by one photon on average in you case.
I installed the newer version of FLAIR v3 from CERN, as you kindly suggested, uninstalling the previous version (v2.2).
I’ve noticed that in this version the RT-viewer box does not appear anymore. I installed FLAIR v3 for 22.0 ubuntu, since the 23.0 version was unavailable
during the installation process, the terminal was compelling about some missing dependencies, like “matplotlib”, “numpy”, “scipy” that I’ve already installed but using anaconda (and they appear writing \verb|pip list|):
Preparing to unpack flair_3.3-1_ubuntu22.04_all.deb ...
Unpacking flair (3.3-1) over (3.3-1) ...
dpkg: dependency problems prevent configuration of flair:
flair depends on flair-geoviewer; however:
Package flair-geoviewer is not configured yet.
flair depends on python3-numpy; however:
Package python3-numpy is not installed.
flair depends on python3-scipy; however:
Package python3-scipy is not installed.
flair depends on python3-dicom; however:
Package python3-dicom is not installed.
flair depends on python3-matplotlib; however:
Package python3-matplotlib is not installed.
flair depends on python3-pil.imagetk; however:
Package python3-pil.imagetk is not installed.
same when instaling the geoviewer package, it has some missing dependencies:
dpkg: dependency problems prevent configuration of flair-geoviewer:
flair-geoviewer depends on flair (= 3.3-1); however:
Package flair is not configured yet.
flair-geoviewer depends on python3-pil.imagetk; however:
Package python3-pil.imagetk is not installed.
while my pip list contains most of the requested ones. it is:
You need to “Import” the USRBIN scoring first to make the RTviewer window available.
The .deb packages depend on the system’s Python packages. dpkg can’t likely recognize the packages supplied by Anaconda. If your Flair is running and you don’t see any other issues, then you can ignore these messages.
