Your pipeline looks a bit strange to me. It seems to confuse the DTI and the T1 data analysis.
I think, you should check the TBSS user guide: http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/TBSS/UserGuide
Cheers,
Auer, Tibor M.D. Ph.D.
MRC Cognition and Brain Sciences Unit
15 Chaucer Road
Cambridge
CB2 7EF
United Kingdom
Phone/Work:
Mobile: +44-(0)7906-86-3837
Mail: [log in to unmask]
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of YX Li
Sent: Wednesday, January 09, 2013 2:53 AM
To: [log in to unmask]
Subject: [FSL] longitudinal analysis on the DTI data
Hi, fsl experts
Happy new year!
I am trying to study the differences in longitudinal changes in two groups (trainees and controls) over the period of three years. Fortunately, the artical of Engivg et al. 2012 (http://onlinelibrary.wiley.com/doi/10.1002/hbm.21370/full) provided the way using TBSS to analysis the longitudinal DTI changes.
Following up the idea of this artical and the past post(i.e. https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind04&L=FSL&P=R55833&1=FSL&9=A&J=on&K=12&X=1B4A95371BC5457B1B&Y=yxin-li%40163.com&d=No+Match%3BMatch%3BMatches&z=4), I give my understanding as bellow:
1. eddy current correction the raw DTI data (time1, time2) and registration within FDT
eddy_correct time1 brain 0
bet brain codif_brain -m -s -f 0.3
dtifit -k brain -o time1_test -m nodif_bran_mask -r *.bvec -b *.bval
the same way was applied to the time2 data.
output: time1_test_FA time2_test_FA
2. For each subject, two-timepoint brain images were resampled into the space halfway between the two
siena time1 time2 -d -B ''-f 0.3''
output: A_halfwayto_B_brain.nii.gz A_halfwayto_B.mat
B_halfwayto_A_brain.nii.gz B_halfwayto_A.mat
3. using FLIRT to linear register the FA images on the halfway space, and average the two halfway space FA maps.
flirt -in time1_test_FA -ref A_halfwayto_B_brain.nii.gz -applyxfm -init A_halfwayto_B,mat -out time1_halfway_FA
flirt -in time2_test_FA -ref B_halfwayto_A_brain.nii.gz -applyxfm -init B_halfwayto_A,mat -out time2_halfway_FA
fslmaths time1_halfway_FA -add time2_halfway_FA -div 2 midspace_FA 4. TBSS analysis
In the analysis of our DTI data, we meet a problem on the step 2. As I know, the outputs of siena (A_halfwayto_B_brain.nii.gz, B_halfwayto_A_brain.nii.gz )were in the same halfway space. Then I want to look at the registration in more detail.
fslview A_halfwayto_B_brain.nii.gz B_halfwayto_A_brain.nii.gz however, we meet a mistake,both images don't have same dimensions.
halfway space data: A_halfwayto_B_brain.nii.gz : voxels: 256*256*128, 1*1*1mm
B_halfwayto_A_brain.nii.gz : voxels: 256*256*127, 1*1*1mm
raw DTI data: time1.nii.gz:voxels: 126*128*64, 2*2*2mm
time2.nii.gz:voxels: 126*128*64, 2*2*2mm
Are there any way to resolve this problem?
I don't know if my understanding of halfway space is correct?
Do you have any comment on this analyses workflow?
Thanks
Yongxin Li
|