RE: [fluka-discuss]: Possible Flair error and stopping power calibration/corrfact/voxel geometry

From: Vasilis Vlachoudis <Vasilis.Vlachoudis_at_cern.ch>
Date: Sun, 16 Oct 2016 07:09:32 +0000

Hi Lars,

indeed it is taken as <= as it is described on the manual. It is only in the name of the provided materials that the = symbol is missing.
In the standard FLUKA release you cannot use the "relative" correction factors when WHAT(1)<0, and since flair is
using that by default (negative logic from the default of fluka) you need to provide all dE/dx correction factors with a negative sign.

The reasoning behind was that in a perfect system you would only need to provide one correction factor in density for all processes
to match the HU. In the "future" release of FLUKA this will be achieved using WHAT(1)=-1. And this is the default in flair.
To have this behavior in the current release you need to provide WHAT(1)=WHAT(2) and in flair you will set the appropriate column of dE/dx as negative.

Vasilis

________________________________
From: owner-fluka-discuss_at_mi.infn.it [owner-fluka-discuss_at_mi.infn.it] on behalf of Lars Fredrik Fjæra [lfj049_at_student.uib.no]
Sent: Friday, October 14, 2016 19:17
To: fluka-discuss_at_fluka.org
Subject: [fluka-discuss]: Possible Flair error and stopping power calibration/corrfact/voxel geometry

Dear FLUKA users,

I have two problems I want to address.

First, is there an error in the Flair Dicom/Voxel tab? The unit range on the far left specifies Hounsfield units "less than" (<), however, when I was using artificial CT-scans, I noticed that Flair applies materials according to units of "less or equal" (<=). As an example, I have the following defined:

< Unit Material Crho_min Crho_max
1000 HU<1000 0.9810 1.0168
1100 HU<1100 0.9852 1.0144

In my CT scan I have only two Hounsfield units, HU=1000 and HU=-1000. Using the HU=1000 as an example, it should, according to Flair, be assigned the material called HU<1100 considering that 999<HU1000<1100. However, the HU of 1000 gets assigned the material HU<1000. This then means that the unit specification in Flair should be (=<). I have also noticed that this was the case in previous versions in Flair referring to for example the image in the manual. I have also attached images of my voxel/geometry setup to further illustrate the problem. The same also happens for my HU=-1000. It gets assigned the material HU<-1000 specified with the unit <-1000, where it really should be assigned the material HU<-1010.

Next problem. I really miss a proper explanation/documentation of the stopping power calibration in the voxel tab. Performing the density calibration (Crho_min and Crho_max) is straight forward and works perfectly. I am however struggling with dE/dx calibration. First off, it is specified that the dE/dx correction are relative corrections. Relative to what? According to the manual, "dE/dx is applied relative to the correction of density (negative WHAT(1) in CORRFACT)". However, I was under the impression that if you perform the calibration in the .vxl file itself, you don't need any separate CORRFACT cards? The manual also states: "CdEdx_rel_min: relative correction factor on dE/dx for the minimum unit in the range CdEdx_rel_max: relative correction factor on dE/dx for the maximum unit in the range". So then I wonder, is the correction applied to the density corrections? I.e. If I have rho = 1, and Crho_max = 1.5 and CdEdx_rel_max = 1.5, the effective density for the stopping power will be 2.25?
Further, from the FLUKA course slides about medical application, there is a sentence that states "FLAIR automatically appends the CORRFACT cards calculated taking into account the calibration curves provided by the user at the end of the .vxl file." So where are these CORRFACT cards?
And to add to all this, I have been playing around with the CdEdx_rel_min and CdEdx_rel_max, just trying to change them to see if I can obtain some difference in the simulated ranges for Bragg peaks. However, I can apply whatever value to CdEdx_rel_min and CdEdx_rel_max, and still obtain exactly the same curves. I've even tried to obtain an error in my simulations by setting the CdEdx_rel_min and CdEdx_rel_max beyond the range [2/3,3/2], however the simulation runs fine. I did the same thing for the Crho_min and max, where I successfully obtained an error. So it seems that FLUKA doesn't even take into account my input to CdEdx_rel_min and CdEdx_rel_max. On the other hand, I am perfectly able to apply corrections using regular CORRFACT cards. Can someone please explain thoroughly how to properly use CdEdx_rel_min and CdEdx_rel_max?

Thank you!

---
Lars Fredrik Fjæra
__________________________________________________________________________
You can manage unsubscription from this mailing list at https://www.fluka.org/fluka.php?id=acc_info
Received on Sun Oct 16 2016 - 10:59:23 CEST

This archive was generated by hypermail 2.3.0 : Sun Oct 16 2016 - 10:59:24 CEST