Dear FSL community,
I am a Bioengineering student using for the first time FSL to carry on a project on DTI data in a neuroscience lab.
I have a study with 35 partecipants, divided into two groups, and several cognitive indices; I decided to use TBSS to investigate possible correlations between those indices and FA.
As a first analysis, I decided to look at my subjects as belonging to one group only, and to look for some regions showing significant correlations between subjects' cognitive score and FA. Therefore I set up my GLM model using one group only, one EV (the index score, demeaned) and two contrasts, to account for positive/negative correlations. My first question is: can I somehow extract the R^2 coefficient from the t-test or p-values that randomise gives me as output?
As a second step, I would like to look at the same subjects but divided into two groups, looking for intergroup differences modulated by my cognitive indices. My second question is therefore if it is correct to use these design matrix and contrasts to look at relevant differences:
EV1 EV2 EV3
1 0 0.31113214
1 0 -0.09452038
1 0 -0.147082564
1 0 0.033198661
1 0 -0.053317603
1 0 0.160626531
1 0 0.013398153
1 0 0.246180027
1 0 -0.025328374
1 0 -0.062327773
1 0 0.016981701
1 0 0.023778081
1 0 0.127093636
1 0 0.136434546
1 0 -0.022854005
0 1 -0.127324152
0 1 -0.048429023
0 1 0.22739822
0 1 0.057931282
0 1 0.041509181
0 1 -0.199771871
0 1 0.077783529
0 1 -0.000248349
0 1 -0.112930714
0 1 0.092357303
0 1 -0.034516769
0 1 -0.116363875
0 1 -0.085483433
0 1 0.230221217
0 1 -0.097082703
0 1 -0.140156675
0 1 -0.257735574
0 1 -0.162054772
0 1 0.119296603
0 1 -0.127792203
/ContrastName1 g1>g2
/ContrastName2 g2>g1
/ContrastName3 pos SwCla
/ContrastName4 neg SwCla
/NumWaves 3
/NumContrasts 4
/PPheights 1.398238e+00 1.398238e+00 4.914719e-01 4.914719e-01
/RequiredEffect 2.230 2.230 2.895 2.895
/Matrix
1.000000e+00 -1.000000e+00 0.000000e+00
-1.000000e+00 1.000000e+00 0.000000e+00
0.000000e+00 0.000000e+00 1.000000e+00
0.000000e+00 0.000000e+00 -1.000000e+00
Every help will be warmly appreciated. I have read so far all the possible user manuals and support pages I could find, but I still am not sure if I am using this tool correctly.
Kindest regards,
Eleonora Mastrorilli
|