Hi everyone,
I want to extract BOLD time-courses from an anatomically defined region (ROI). My data was pre-preprocessed using FMRIPREP. While there are of course various ways to achieve this, some people suggest that, e.g.,
[…] the best practice is to define such ROIs for each subject based on their own anatomy, since there can be substantial variability between individuals in macroscopic anatomy" (Poldrack, 2007, NeuroImage)
If I understand correctly, Freesurfer performs automated anatomical labeling of individual anatomical images as part of the FMRIPREP workflow. I assume that the corresponding labels are stored under /freesurfer/{subject-label}/label/, with labels like rh.V1_exvivo.label.
Basically, my question is whether one can combine several labels of choice to create an individual ROI that matches the BOLD data in T1w-space?
Here is what I tried so far:
I already found that the Freesurfer command mri_mergelabels can be used to merge labels.
Next, I thought one could use Freesurfer’s mri_vol2roi to create a binarized brain mask that fits the subject’s BOLD data (in T1w-space). The typical command looks like this:
mri_vol2roi --label your.label --srcvol f --srcreg register.dat \
--finalmskvol labelbinmask --roiavg /tmp/not.wanted.dat
However, I was not sure which FMRIPREP would be the correct one to use for the --srcreg register.dat argument. Any ideas?
A similar topic was already discussed here but I could not really translate the proposed solutions to my case.
Thanks in advance for your help!