I don't think FSL has tools to do this sort of thing (i.e. making ROIs at
the grey-white boundary). You could do it in Caret though. If you have
your activation map on the surface you could threshold that and then map the
data from the white matter surface to the volume and binarize it once it was
in the volume with fslmaths.
There is also probably a FreeSurfer way.
Peace,
Matt.
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf
Of Robert Schulz
Sent: Thursday, September 30, 2010 3:24 AM
To: [log in to unmask]
Subject: [FSL] DTI and fMRI activities: creating seed masks using 'blobs'
Hi FSLers,
1) currently I am trying to figure out how to set the seeding and target
masks for a group of subjects for DTI connectivity analyses in combination
with fMRI activity maps. What is your suggestion for creating the masks when
variability of the step should be minimized?
2) In fact if possible I don't want to draw the masks for every single
subject by hand in the diffspace. What I think of is creating an interface
mask at the grey-white matter boundary adjacent to a fmri motor spot and
extract the 100 nearest mask-voxels to this activation peak. Since this
activition peak (x,y,z) is in derived in mni standard space (using SPM), I
thought to create this interface mask for each ROI in the FSL MNI template.
Does this sound reasonable to you? Will FLIRT/FNIRT work to create proper
stand2diff transformations for the semi-automated seeding masks?
3) Generally, what are your experiences in creating masks for specific
cortical functional areas (like PMd) minimizing the variability and
investigator dependency? Is there a tool for automated mask creation
targeting specific cortical region? What should I use? What would be the
pipeline?
As new to FSL I thank you so much for your comments and suggestions!
Thanks a lot
Kind regards
Robert
------------------------------------
Robert Schulz
University Medical Center Hamburg-Eppendorf
Department of Neurology
Brain Imaging and NeuroStimulation lab (BINS)
Martinistraße 52
D - 20246 Hamburg
email. [log in to unmask]
|