Hello,

I am trying to calculate a linear regression between a continuous variable and grey matter in each voxel of MRI data, with TFCE, using fsl randomise.

The thing is, my outputs do not match what I obtain with SPM or FEAT. Actually, I am not even sure that randomised offers the possibility to do it, I don't find the documentation very clear.

My code is : 

randomise -i spmerge.nii -o randomise -d design.txt -t contrast.txt -m maskSG.nii -T

spmerge.nii is the 4D image which I obtained by merging my 3D smoothed grey matter MRI, with SPM

design.txt is a vest file containing the value of my variable of interest, in column (1 row of 1 value per subject)

contrast.txt is a vest file containing "0 1"

maskSG.nii is my grey matter mask

Do you know if voxelwise correlation are actually possible with randomise, and if so how I should change my code for it to work ?

Thank you in advance !