its possible to do 1). see the 'ROI Analysis' and 'Group Analysis' tutorials under:

http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial

although doing this seems like 'double-dipping', or 'circular analysis'.

2) is possible, as it seems like an extension of 1).

n.

On Sun, 2010-01-31 at 14:20 +0900, Manabu Kubota wrote:

Thank you very much for replying my question.

I understand how to define ROI, and how to extract data.

I have a further question.

Since relatively wide regions are found as significant thickness reductions in the patient group, I want to do either

1)make a mask where the patient group showed significant reduction, and using it for further correlational analyses with clinical variables, to define a "peak vertex" in each region (in the voxel-based morphometry study in SPM, it is easy to do correlational group study in the masked region only (not entire brain), I'm not sure if we can do similar way in freesurfer)

or

2)define ROIs using each region of the significant reduction (I understand how to do this now), and if the ROI is wide, divide one ROI into several region (suppose one ROI include both Orbito Frontal and Superior Frontal, devide it according to the Desikan template), and search correlation with clinical variables, using SPSS

I compared cortical thickness between two groups (a control group and a patient group), and found several areas with significant thickness reduction in patient group.

Is it possible to do 1) and 2) with qdec or tksurfer? Or can I do these from command line?

Thanks,

Manabu Kubota kubotajp@kuhp.kyoto-u.ac.jp Graduate school of medicine, Kyoto University

2010/1/29 Nick Schmansky nicks@nmr.mgh.harvard.edu:

...

the qdec tutorial describes how to draw a label on a region, and then map that label (roi) back to each subject. from there, aparcstats2table can be used to gather that roi data into a table for import into a stats program where you perform your analysis.

n.

On Wed, 2010-01-27 at 08:04 +0900, Manabu Kubota wrote:

...

Dear all,

Hi, I am a new user of freesurfer, and have a questions on group analysis with qdec.

I compared average thickness between two groups (a control group and a patient group).

What I am trying to do next is to make a mask where the patient group showed reduction of cortical thickness (for example, p<0.05, FDR), then examine the correlation between clinical scales and cortical thickness in the masked regions, by doing group analysis in the regions (not entire brain).

I searched past archives, but couldn't find one. Does anyone know how to do this with qdec ?

Thanks for your help,

Kind regards

Manabu Kubota kubotajp@kuhp.kyoto-u.ac.jp Kyoto University _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer