mc-z is a simulation that replaces the data with smoothed gaussian noise. permutation uses your data and changes the order of the rows in the design matrix. If the noise in your data set is Gaussian, they will give the same result (in theory:). If the noise is not gaussian, then the mc-z will give incorrect p-values. With permutation (at least our implementation), the design matrix is supposed to be orthogonal, which is a limitation. doug

On 02/07/2013 03:55 AM, Gabriel Gonzalez Escamilla wrote:

Thanks you for your quick answer Doug,

Just one clarification, in the correction for multiple comparisons, which is the main difference between the mc-z and the permutation test? rather than making simultaions or working on original data.

And Why is 5000 runs recommended to mc-z and 10000 for permutation testing?

Thanks, Gabriel.

El 07/02/13, *Douglas N Greve * greve@nmr.mgh.harvard.edu escribió:

...

On 02/06/2013 12:49 PM, Gabriel Gonzalez Escamilla wrote:

...

Dear Doug and FreeSurfer experts,

Thnak you so much for all your lasts responses!.

I did ran succesfully all the pre-processing for the interhemispheric comparisons on FS. Now I have a couple questions:

given the command for interhemispheric comparisons: mri_glmfit --y lh.lh-rh.thickness.sm05.mgh --glmdir glm.lh.lh-rh.thickness.sm05 --osgm --surf fsaverage_sym lh;

A) This read all subjects to construct X and C as one-sample group mean, this means that as result I get the difference across hemispheres and across my subjects in a population, right?

Correct.

...

B) But, what if I have two groups? controls and patients, I'm wondering if is it possible to examine regional differences of thickness asymmetries between those two groups? or more?

This just becomes a standard two group analysis as described in the tutorials. You create an FSGD file with two classes and list your input subjects.

...

C) If I want to examine effects of group or sex, while controlling for age and other factors how do I introduce them into the glm model? If so, do I set the contrasts as normally for mri_glmfit? I mean as any other thickness study? or there is any other thing that I should do?

You do not need to do anything different than with a normal thickness study.

...

The command for multiple comparisons correction is: mri_glmfit-sim --glmdir glm.lh.lh-rh.thickness.sm05 --cwpvalthresh .5 Which performs only cluster-wise corrections,

D) is there any other way to correct for multiple comparisons, something like permutation tests or anything else?

Yes, you can run something like mri_glmfit-sim --glmdir glm.lh.lh-rh.thickness.sm05 --cwpvalthresh .5 --sim perm 10000 2 csdperm --sim-sign abs You should run mri_glmfit-sim --help to see what these options mean as you may need to change the voxelwise threshold (the "2" value)

doug

...

Many thanks in advanced, Gabriel

-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422

Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/

---

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

The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.

--

PhD. student Gabriel González-Escamilla Laboratory of Functional Neuroscience Department of Physiology, Anatomy, and Cell Biology University Pablo de Olavide Ctra. de Utrera, Km.1 41013 - Seville

• Spain -

Email: ggonesc@upo.es http://www.upo.es/neuroaging/es/