Thanks Doug.

To check if I get the same results (as obtained using FDR) using clusterwise correction for multiple comparison I ran a Z Monte Carlo simulation on my data with the following command:

'mri_glmfit-sim --glmdir $SUBJECTS_DIR/sex_diff_paired_all/lh.paired-sexdiff --sim mc-z 5000 2 mc-z.abs --sim-sign abs'

In the contrast subdirectory I get the following files: mc-z.abs.sig.voxel.mgh mc-z.abs.pdf.dat mc-z.abs.sig.cluster.mgh mc-z.abs.sig.cluster.summary mc-z.abs.sig.ocn.lut mc-z.abs.sig.ocn.mgh

but the cluster annotation file, which is necessary for viewing the results is not in this subdirectory.

Do you have any idea what went wrong?

Thanks.

Anouk

-----Oorspronkelijk bericht----- Van: Douglas N Greve [mailto:greve@nmr.mgh.harvard.edu] Verzonden: maandag 29 augustus 2011 18:58 Aan: Braber, A. den CC: 'freesurfer@nmr.mgh.harvard.edu' Onderwerp: Re: [Freesurfer] Paired analysis results into table?

Hi Anouk, it you look at the voxel-wise threshold, you can then run mri_surfcluster using that threshold to give you a table of clusters. You should be careful how you interpret the FDR results. Technically, you do not get "clusters" from FDR, and the clusters you get from the table are not "corrected" in the way that they would be corrected with gaussian field theory.

doug

Braber, A. den wrote:

Hi,

I just performed a paired analysis on my twin data. I am interested in the difference in cortical thickness between males and females, and have to run a paired analysis since the males and females in my sample are related (because they are twins). No covariate (such as age) has been included in the mri_glmfit yet, so I only have 1 contrast (mean.mtx = 1, meaning in my case males greater than females). After running the mri_glmfit I viewed my results using

' tksurfer fsaverage lh inflated -annot aparc.annot -fthresh 2 -overlay $SUBJECTS_DIR/ sig.mgh'. Thereafter I corrected my results using FDR-correction, under View->Configure->Overlay->set threshold using FDR, rate 0.05 (so I did not perform a clusterwise correction for multiple comparison). Tksurfer than gives me a really nice overview of the significant results (FDR-corrected) of my paired analysis. I wondered whether there is a tool to export these results into a table (just like it's possible with aparc2table), so I get an overview of, for example, T-statistics, p-value and vertex number, for the observed significant clusters (or vertex with the highest T-value within an observed cluster). Is this possible? I have searched for a command to do this, but unfortunately I couldn't find it. Could you please help me?

Thanks in advance.

Anouk den Braber

---

---

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