Dear FSL users,
I could really appreciate your expertise on building up the right contrasts for my analysis as unfortunately the answers I have found on the mailing list have not been extremely helpful.
I want to compare two pairs of control-versus-patient groups to examine whether the Control Group 1 relates to Patient group 1 the same way as Control group 2 relates to Patient group 2 [in other words C1>P1 ~ C2>P2]. At this stage I would only like to adjust for age differences.
I would like to covary for age and gender as well. I have so far simply included these as 2 extra EVs at the end (without demeaning, I get the impression that in this case it would not make much difference, if wrong please correct me).
randomise -i all_FA_skeletonised.nii.gz -o tbss -m mean_FA_skeleton_mask.nii.gz -d design.mat -t design.con -f design.fts -n 5000 --T2
/NumWaves 6
/NumPoints 112
/PPheights 1 1 1 1 1 1
/Matrix
1 0 0 0 24 0
1 0 0 0 38 0
1 0 0 0 49 1
1 0 0 0 26 1
1 0 0 0 39 1
1 0 0 0 34 0
.....
where EV1=C1; EV2=P1; EV3=P2; EV4=C2; EV5=age; EV6=gender
/NumWaves 6
/NumContrasts 4
/PPheights 1 1 1 1
/Matrix
1 -1 0 0 0 0
-1 1 0 0 0 0
0 0 -1 1 0 0
0 0 1 -1 0 0
where:
#/C1 C1>P1
#/C4 C2>P2
#/adjusted for age and gender
/NumWaves 4
/NumContrasts 1
/PPheights 1 1 1 1
/Matrix
1 0 1 0
I presume that with this given design.fts I am looking at non-directional difference across all 4 groups. However what would be the best way to compare the first difference (C1-P1) with the second (C2-P2)? I really can`t see a way to execute this analysis when there are 4 independent groups.
Any suggestions/help would be greatly appreciated!!!!
Regards,
Linn Mittle
|