Dear FSL users & experts,
Apologies for resurrecting this old post, but after trying its suggestions out, I'm still getting suboptimal results so thought I'd check to see if anyone can think of a way to improve on my efforts!
I am trying to create a seed mask to use in Probtrackx and have done the following:
1) Used the atlas tool to create a unilateral M1 mask using the standard 1mm brain as template.
2) Created a transformation matrix (to go from diffusion space to standard space) and inverted it with the following:
flirt -in <diffusion data> -ref <MNI 1mm brain> -omat <diff2MNI.mat>
convert_xfm -omat <MNI2diff.mat> -inverse <diff2MNI.mat>
For the input diffusion data, I have tried using: the reconstructed dti_FA map (from running DTIFIT), the whole diffusion dataset image, and the nodif.nii.gz (I think this is the b0 image from the diffusion dataset).
3) I then used this diff2MNI.mat to transform my standard-space mask to subject diffusion space using:
flirt -in <mask> -applyxfm -init <diff2MNI.mat> -out <maskv2> -paddingsize 0.0 nearestneighbour -ref <nodif.nii.gz>
The problem is that the resulting mask, when viewing it on nodif.nii.gz, is a little displaced posteriorly compared to where it should be.
Am I making some basic mistake in how I'm generating the matrix? or doing the transformation?
Many thanks in advance for your help!
Kind regards,
Lucia
|