Well, to be short - I would >not generally< treat the tumor as an exclusion or termination criterion.
However, I would consider a lot of things:
- do the tracts you are trying to reconstruct pass through the tumor?
- is the tumor accompanied by perifocal edema?
- do the tracts pass through the perifocal edema?
- is the edema less or more after the surgery?
- how much time has passed after the surgery?
- does the patient get steroids? does that change pre/post?
are we talking a neural or glial tumor or metastasis, or all mixed?
... (tbc)
So yes, this is a methodological question and the answer will also depend what you are after. if you are after remote connectivities distant to the tumor and the potential perifocal edema, you may indeed consider to exclude the tumor / immediately adjacent peritumoral areas.
hth,
andreas
Am 11.09.23, 17:07 schrieb "FSL - FMRIB's Software Library im Auftrag von Pia Ritter" <[log in to unmask] im Auftrag von [log in to unmask]>:
Hi Andreas,
thank you for your answer!
we are interested in looking at how structural connectivity changes in tumor patients from pre- to postoperatively (specifically whether, and if so where, there is reduced connectivity). For this we want to create structural connectivity fingerprints where we look at the connectivity of a seed to different targets. The seeds and targets are cortical areas in the ipsilesional hemisphere, and we have created ROIs at the GM/WM border. The target ROIs are broadly distributed across the cortex in frontal, temporal but also parietal areas. The seed ROI is the IFG. In principle, we are interested in the pattern of connections of the IFG with other cortical ROI - HOWEVER we are NOT interested in group-level analyses, but rather want to look at how structural connectivity changes from pre- to post-OP in each patient individually!!
Brief overview of the analysis: We do this by seeding from the seed with probtrackx2, without specifying targets. I then apply the logarithm to the output file, fdt_paths.nii.gz, and then normalize all voxels by dividing them by the highest voxel value in the brain to obtain values between 0 and 1. Following this, I overlay the targets on the modified file (fdt_paths_log_div.nii.gz) to extract the average connectivity value, normalized with respect to the strongest connectivity in the brain, for the voxels within each target.
As we are not interested in group-level analysis, our cohort is not very homogeneous in terms of tumor type (i.e. high- and low-grade tumours are included) and also in terms of the tumor localisation - but all tumors are located in or near speech-relevant areas! It is therefore expected that the structural connectivity is somehow affected by the tumor, however, as we are looking at subject-specific changes in structural connectivity I was wondering whether it makes sense to take the tumor into account somehow.
The longer I write this mail, the more I realise that this is probably more a methodological question than something to do with probtrackx2... but I would still be grateful for any input, as I am still quite new to the field of structural connectivity analysis.
Thanks for your help,
Pia
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=FSL&A=1
This message was issued to members of www.jiscmail.ac.uk/FSL, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=FSL&A=1
This message was issued to members of www.jiscmail.ac.uk/FSL, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
|