Nirstorm/nirstorm

Issue with segmentation

GuyRens opened this issue · 4 comments

GuyRens commented

Dear community,

I'm trying to compute the fluency model relying on default anatomicals.
In a recent study, we collected data by placing optodes following the 10-20 system which uses the ICBM-152 template for optode positioning.

However, I am stuck on the problem that I do not have the fMRI segmentation. See error in image
compute_fluence_issue.

In addition, when I try to calculate this segmentation in the anatomy screen using cat12:
right-click MRI: ICBM152 > cat12 MRI segmentation, I get the following error:

cat12_issue

Does anyone know how to solve this issue of not getting the segmentation for the default anatomicals (such as Collin, ICBM-152))?

Many thanks in advance

Guy Rens

zhengchencai commented

Hi Guy,

The segmentation will be used to perform the Monte Calo simulation to solve the forward problem which "links" the cortical space and channel space. The file is imported as .nift format however calculated by any segmentation method one prefers. It has to have 5 tissue types (i.e. skin skull, CSF, white matter and gray matter) labelled as 1, 2, 3, 5, 4, and 0 as background. Brainstorm build-in segmentation may not able to manage this mission since it's not designed for NIRS. You could use FSL or SPM or Freesurfer. There are some tricks to do this in Brainstorm feel free to contact us.

Thanks,
Zhengchen

GuyRens commented

Hi Zhengchen,

Thank you for this information. I already noticed that I need the segmented file for solving the forward problem.

Unfortunately, I do not seem to find a solution to achieve this segmentation based on the atlas files that are on the brainstorm website or by right clicking on 'default anatomy' in the brainstorm MATLAB GUI > click on 'use template' > select one of the templates.

I would be very grateful if you could teach me this segmentation, preferably in brainstorm, but by any other means if not possible in brainstorm.

If you are interested in helping me, please let me know what is most convenient for you.
I'm more than happy to do this over comments, zoom, e-mail, etc.

Thank you in advance

Guy

zhengchencai commented

Hi Guy,

For performing segmentation really it depends on which software you'd prefer. I encourage you to start with FSL or SPM, they both provide 5 types of tissues. They also have detailed documentation and the community to discuss certain issues.

Thanks,
Zhengchen

GuyRens commented

Dear Zhengchen,

Thank you for your help.
I tracked down my problem to a potential bug in the brainstorm software itself.

I reported this on their forum as it is less relevant here. If they could address this issue, it should be fixed for nirstorm as well.

Again, many thanks

Guy