Hi,

 

I have performed a linear mixed model analysis on surface data in AFNI using SUMA, resulting in each vertex having a p value. I would like to perform a multiple comparison correction using freesurfer, as an FDR approach in AFNI is to stringent (individual vertex statistical thresholds are low). I would like to run a monte carlo simulation using mri_glmfit on a similar surface with similar parameters to extract a cluster threshold. I am not used to using freesurfer, so bear with me, please.

 

1) I have created a fsgd file defining the structure of the analysis used. This seems to be the tricky part, as I don't know how to exactly specify the model used. Maybe it suffices to define the data structure.

There is two groups (patients, controls) and three timepoints. I want to correct for the group between group difference (contrast 1 -1), but don't know how to specify multiple timepoints per subject.

 

2) I have smoothed the data with fwhm 3mm in afni. How can I perform the simulation using mri_glmfit-sim and how can I do this without doing the analysis itself in freesurfer?

 

I would appreciate your advice!

Casper

--

Casper A.M.M. van Oers, MD

Department of Neurology, C03.236
University Medical Centre
PO Box 85500
3508 GA Utrecht
the Netherlands

T: +31 30 2503184
F: +31 30 2542100 (Dept. of neurology)