Dear FreeSurfer,

I would like to do the long. two stage model (https://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalTwoStageModel)

Study design: 2 groups (one control group, one patient group) with gender and age as covariates. Each subject was MRI scanned at baseline and 6 weeks follow-up. I would like to test the difference in structural measures between groups over time.

My problem

1. Is the process and setup below correct?

2. When i pass the above command under 3) i get an error where it says:

FSGDF Format Error: file = /Volumes/External/P1_FU/GroupDiffLhThickness.fsgd, tag=Input

Is there anyone that can help?

I have followed the tutorial:

FIRST STAGE: Preparing the Data - QCACHE is done.
SECOND STAGE: Option mri_glmfit:

Then i would like to test the difference between groups, and this is were i have a hard time following the tutorial.

I looked at the following guide: https://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/GroupAnalysis concerning Group Analysis using mri_glmfit and i understand that i need to use the command below creating my own fsgd file.

1) Since i have a longitudinal setup, i converted it to a cross-sectional steup as you would do if you used the QDEC option. My fsgd file looks as the one below. 1 = patients and 0 = controls.

GroupDescriptorFile 1
Title GroupDiff_Lh_Thickness
Class 1
Class 0
Variables age gender
Input b101_base 1.0 24.0 1.0
Input b103_base 1.0 41.0 0.0
Input b105_base 1.0 22.0 1.0
Input b206_base 0.0 25.0 1.0
Input b207_base 0.0 23.0 0.0

2) To test group difference between groups regressing out the effect of age and gender i have made a .mtx file with the following input:

1 -1 0 0 0 0

3) Therenext, i ran the following but something is wrong with my fsgd file:

mri_glmfit --y lh.thickness-spc.stack.fwhm15.mgh --fsgd GroupDiffLhThickness.fsgd dods --C group.diff.mtx --surf fsaverage lh --cortex --glmdir lh_thickness_spc_fwhm15_groupdiff

I hoper there is someone who can help.

Best wishes

Kasper Jessen