Thanks, but I got the following error message:

RROR: matrix is ill-conditioned or badly scaled, condno = 1e+08

--------------------------------

Possible problem with experimental design:

Check for duplicate entries and/or lack of range of

continuous variables within a class.

If you seek help with this problem, make sure to send:

1. Your command line:

mri_glmfit.bin --y rh.t1t2ratio_30_fmed_3sm_decurv.fsavg.mgh --fsgd age_gender.fsgd --pvr rh.thickness_30.fsavg.mgh --C group.diff.mtx --surf fsaverage rh --cortex --glmdir rh_allbrain_t1t2_group_thickness_age_gender.glmdir

2. The FSGD file (if using one)

3. And the design matrix above

Attempting to diagnose further

SumSq: Min=0.000000 (col 5), Max=146.140350 (col 3)

The scale is much different between columns 5 and 3, you may want to

normalize by subtracting the mean and dividing by the standard deviation.

Column 5, all values are 0

Column 6, all values are 0

Columns 5 and 6 are the same

Il giorno dom 6 gen 2019 alle ore 20:18 Greve, Douglas N.,Ph.D. <DGREVE@mgh.harvard.edu> ha scritto:

yes, that is the right contrast

On 1/5/19 7:30 AM, Raffaele Dubbioso wrote:
External Email - Use Caution

Thanks a lot.
Sorry if I bother you again:

If I want to remove thickness, age and gender as nuisance variables;

My FSGD file should be:

GroupDescriptorFile 1

Title Func

Class HC

Class CMT

Variables Age gender

Input fs_VF HC 46 Female

Input fs_RF HC 24 Female

.......

My group.diff.mtx should be:

1 -1 0 0 0 0 0

Is it ok?

Thanks a lot

Raffaele

Il giorno gio 3 gen 2019 alle ore 20:11 Greve, Douglas N.,Ph.D. <DGREVE@mgh.harvard.edu> ha scritto:

The contrast is not correct for your design. Your design matrix will
have 3 columns, one for HC, one for CMT, and one for
rh.thickness_30.fsavg.mgh. Ie, it automatcailly interprets PVRs in a
DOSS kind of way. If you want to test for a difference in
thickness-meylin slopes between the two groups, you will need to have
two PVRs, one where the value is the thickness for HC subjects and 0 for
CMT subjects, the other the reverse, then set your contrast to 0 0 1 -1.
If you just want to remove thickness as a nuisance, then you can stick
with what you have and just use a contrast of 1 -1 0

On 12/25/18 5:05 AM, Raffaele Dubbioso wrote:
>
> External Email - Use Caution
>
> Dear Freesurfer team,
>
> I'm running a group analysis where I'd like to compare myelination
> between two groups regressing out cortical thickness.
> Unfortunately I get always the same error message:
> dimension mismatch between X and contrast group.diff.mtx X has 3
> cols, C has 4 cols. Is my group.mtx correct?
>
> My command line is:
> mri_glmfit --y rh.t1t2ratio_30_fmed_3sm_decurv.fsavg.mgh --fsgd
> thickness.fsgd --pvr rh.thickness_30.fsavg.mgh --C group.diff.mtx --C
> group-x-thick --Cg1g2.intercept.mts --C g1g2.thick.mtx --surf
> fsaverage rh --cortex --glmdir rh_allbrain_t1t2_group_thickness.glmdir
>
> My group_diff.mtx: 1 -1 0 0
>
> This is my FSGD FILE:
> GroupDescriptorFile 1
> Title Myelin
> Class HC
> Class CMT
> Inputfs_VFHC
> ........
>
>
> Thank you very much for your help
>
> Best regards,
> Raffaele
>
> _______________________________________________
> Freesurfer mailing list
> Freesurfer@nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

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

--

_______________________________________________
Raffaele Dubbioso, MD, PhD
University Federico II of Naples

Department of Neurosciences, Reproductive Sciences

and Odontostomatology.
Via Sergio Pansini, 5
80131 Napoli, ITALY
phone: +39-0817464579
fax: +39-0817462667
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer