On 10/29/2021 10:39 PM, Zeng, Qi wrote:

        External Email - Use Caution        

Hi Experts, 

After completing long_mris_slopes to get the rate of change and then concatenating them into one.mgh file for group analysis (mris_preproc). I am finally getting to mris_glmfit. Got a couple of questions along the way.
1). if I conducted long_mris_slopes, and got for example, lh.long.thickness-rate.fwhm10.fsaverage.mgh. Do I need to do a surface smoothing again (mri_surf2surf) before mris_glmfit?
No
2). I construct covariates in FSGD besides the group contrast that I am interested in. How do I add covariates into the glmfit model to control for the covariates as I contrast groups in terms of the rate of changing (CV vs. NCV groups) . Do I add them into --C as  (1, -1, 0(age), 0(gender), 0(race)). 
mri_glmfit 
        --y {$hemi}.{$meas}-rate.{$smoothness}.mgh 
        --fsgd FSGD/{$study}.fsgd doss
        --C Contrasts/CV-NCV.mtx    (1, -1)
        --C Contrasts/NCV-CV.mtx     (-1, 1)
        --surf fsaverage {$hemi}                                
        --cortex                                               
        --glmdir {$hemi}.{$meas}-rate.{$study}.{$smoothness}.glmdir

Not sure what you mean here. Your FSGD file will translate into a design matrix. Each column in the design matrix means something (eg, group membership, covariate of a group). You need to create a contrast matrix that embodies your hypothesis. There needs to be an item in the contrast matrix for each column in the design matrix. See https://surfer.nmr.mgh.harvard.edu/fswiki/FsgdExamples for more info

Best, 
Zeno


--



_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer