Dear FSL experts,
After unpaired 2 group (12 control vs 12 patient, age matched) TBSS analysis, following standard pipeline and using TFCE option on randomise, I got widespread differences of FA even with corrected p=0.01.
Design matrix
Group EV1 EV2
Input1 1 1 0
Input2 1 1 0
......
Input13 2 0 1
Input14 2 0 1
......
My contrasts
groupA groupB
C1 1 -1
C2 -1 1
If my design matrix is correct, I should assume that it is a global effect, shouldn't I?
Next, although my groups are aged matched, I decided to control for age and gender. My design matrix was following:
Group EV1 EV2 EV3(age) EV4(gender)
Input1 1 1 0 1,16 -1.16
Input2 1 1 0 -1,26 0.83
......
Input13 1 0 1 -0.65 0.83
Input14 1 0 1 0.23 0.83
......
My contrasts
groupA groupB age gender
C1 1 -1 0 0
C2 -1 1 0 0
Hope, that's ok. I used grand mean for age and gender.
However, I have confusion for my next question, whether there is a correlation (negative or positive) between FA and clinical scores (SPRS)? The last is a spastic paraplegia rating scale and ranges from 0 (no spasticity) to 52 (max spasticity).
Do I need to enter raw SPRS scores or I need to demean, if so how? Is overall my design matrix correct for answering my question?
Group EV1 EV2 EV3(age) EV4(gender) SPRS
Input1 1 1 0 1,16 -1.16 0
Input2 1 1 0 -1,26 0.83 0
......
Input13 1 0 1 -0.65 0.83 ?
Input14 1 0 1 0.23 0.83 ?
......
groupA groupB age gender SPRS
C1 1 -1 0 0 0
C2 -1 1 0 0 0
C3 0 0 0 0 1
C4 0 0 0 0 -1
Thank you very much
Gayane
|