There are several approaches, e.g. :
fslmerge -t big4d data1 data2 #merge the averages into a single 4D file
bash eddy_correct big4d big4d_corr 0 #motion correct all data to first
volume
bash -c "fslroi big4d_corr av1_corr 0 40"
bash -c "fslroi big4d_corr av2_corr 40 40"
fslmaths av1_corr -add av2_corr -div 2 data #average motion corrected
data
Which averages the data before you run DTIFIT.
Or simply doing the first two steps of the above and concatenating your
bvals and bvecs files accordingly and then running DTIFIT. This method may
make your bedpostx preprocessing time take longer; however, but it allows
you do such things as correcting gradient table for subject motion, if you
so desire to do that.
Peace,
Matt.
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf
Of Moran Artzi
Sent: Tuesday, April 21, 2009 2:27 AM
To: [log in to unmask]
Subject: [FSL] dtifit with multiple repetition
Hi,
We have DTI data with no of repetition (2-4 NEX) that we would like to run
using the dtifit tool.
So for data with 2 NEX and 15 directions we have 32 files (B0*2 + 15dir*2).
1. How should I run the "eddy current" in order to align al files to the
second
B0?
2. for better SNR- How should I run the "dtifit" tool? Do I need to sum the
4D
(data.nii.gz) files/multiply or edit the bvecs & bvals files or otherwise?
Many thanks
Moran
|