Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

NaN values in physicalDose of resultGUI [BUG] #611

Open
zperko opened this issue Apr 20, 2023 · 2 comments
Open

NaN values in physicalDose of resultGUI [BUG] #611

zperko opened this issue Apr 20, 2023 · 2 comments
Labels

Comments

@zperko
Copy link

zperko commented Apr 20, 2023

Describe the bug
In some cases the returned physicalDose array contains NaN values, while the none of the beam physical dose arrays (physicalDose_BeamX) have any NaNs in them. Seems like this mainly happens along the edges of the CT (e.g., resultsGUI.physicalDose(:,:,end) or (:,end,:), so it could be a loop end index being incorrect.

To Reproduce
Bit difficult without providing the patient data, but we have seen this for multiple brain cases.

Expected behavior
Should not have NaN values in calculated dose arrays, and the beamwise dose values should correspond to the total dose array.
See the attached screenshot of a normal looking dose and DVHs, but with NaNs in the physical dose.
Screenshots
image
image

Desktop (please complete the following information):

  • OS: UNIX
  • Environment & Version Matlab 2021b
  • matRad Version 2.10.1

Additional context

@zperko zperko added the bug label Apr 20, 2023
@zperko
Copy link
Author

zperko commented Apr 20, 2023

Actually, there are also slices only partially filled with NaNs, so probably not a looping issue. Still though these seem to be always close to the edges of the CT.
image

@wahln
Copy link
Contributor

wahln commented Apr 21, 2023

Do you know what rsp / HU values are present during dose calculation in these voxels? Might be a little annoying to debug due to the grid's usually being different and the resampling involved. I guess the NaN's are somewhere created in interpolation when a value lies out of range, so this could either be the 3D grid resampling or Hounsfield Look-Up.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests

2 participants