Hello, I would like to first people for the help with the fieldmapping
suggestions. It worked perfectly.
However, I am now having some problems with using FLIRT to align my mean EPI
to my MP-RAGE (SPGR) anatomical image. For some participants, it works just
fine, but for many others, the resulting image either places the prefrontal
cortex in the cerebellum and the occipital in the parietal, or the image is
completely upside down in the axial plane. I am assuming that I must be
doing something wrong (it works right for only about 25% of participants).
I have used several variants of the code below.
flirt -ref anat -in meanEPIbet -omat my_affine_transf.mat -out meanFLIRT_epi
-dof 6
for example:
flirt -ref anat -in meanEPIbet -omat my_affine_transf.mat -out meanFLIRT_epi
-dof 6 -finesearch 25 -cost mutualinfo -searchcost mutualinfo
or
flirt -ref anat -in meanEPIbet -omat my_affine_transf.mat -out meanFLIRT_epi
-dof 6 -cost mutualinfo -searchcost mutualinfo
for the anatomical image, I have tried skull stripping using bet2, and
AFNI's 3dSkullStrip, and I have tried betting the meanEPI (and not as well).
bet2 s002a1001.hdr anat
bet2 meanEPI meanEPIbet
or
3dSkullStrip -input s002a1001.hdr -prefix anat.nii.gz
3dSkullStrip -input meanEPI.nii.gz -prefix meanEPIbet.nii.gz
or
the ones above, just for the anatomical image
By going through several iterations, I can usually find one that aligns the
EPI to the structural, but this is taking about 30-40 minutes per
participant randomly playing with parameters.
We are using a Trio Tim system, and not really doing anything special with
the data collection.
If anyone has any suggestions, that would be great.
Best,
Wil
|