Dear Freesurfer experts,
I have problems concerning the results of a MC simulation. When I checked the results with Qdec I discovered a matching problem for the right hemisphere. It seems like the used command runs the correct analyses for the right hemisphere, but fails projecting it properly to that one. It looks like freesurfer uses atlas and coordinates of the left hemisphere for the right hemisphere. I think the problem starts during MC simulation. Here's a "copy paste" entry of the *.log file, where I think freesurfer clearly mixes up hemisphere information:
version $Id: mri_surfcluster.c,v 1.39.2.3 2009/01/15 01:06:01 greve Exp $ hemi = lh srcid = /home/hiwibp14/smb4k/CORTICAL/fmrt/Studie43/freesurfer/qdec/Untitled/rh-Avg-Intercept-thickness/sig.mgh
For me obviously in the simulation process the programm mixes up information for right and left hemisphere during the analyses. This happens several times during analyses but only for analyses concerning the right hemisphere (see the complete attached log file). Is there any possibility to specify the command, adressing this problem like a special flag or something? Or is it possible to run the right hemisphere again separately?
To run the simulation I used the following command line:
mri_glmfit-sim \ --glmdir /home/hiwibp14/smb4k/CORTICAL/fmrt/Studie43/freesurfer/qdec/Untitled \ --sim mc-z 5000 1.3 mc-z.abs.1.3 \ --sim-sign abs \ --overwrite
Thanks for your help and best regards,
Johanna Feiler ___________________________________________________________ WEB.DE DSL ab 19,99 Euro/Monat. Bis zu 150,- Euro Startguthaben und 50,- Euro Geldprämie inklusive! https://freundschaftswerbung.web.de
Johanna,
Can you try out a new version of qdec that incorporates an implementation of MC simulation that runs instantly?
For the linux x86_64 platform, you can get qdec.bin from here:
ftp://surfer.nmr.mgh.harvard.edu/pub/dist/freesurfer/misc/linux-centos4_x86_64/
copy qdec.bin to your $FREESURFER_HOME/bin, and type:
chmod a+x $FREESURFER_HOME/bin/qdec.bin
you will also need to get a set of pre-run simulation data. download the file mult-comp-cor.tar.gz from here:
ftp://surfer.nmr.mgh.harvard.edu/pub/dist/freesurfer/misc/
and copy it to $FREESURFER_HOME/average
then type:
cd $FREESURFER_HOME/average tar zxvf mult-comp-cor.tar.gz
Then when you run qdec, after an analysis is complete and a contrast result is selected, the MC Simulation 'Run' button will perform the simulation and load the result.
When qdec starts, it will give some error about missing icons. Just click past the error (or copy the files icon_curv_bw.gif and icon_curv_bw.tif to $FREESURFER_HOME/lib/images).
This new qdec will appear in the upcoming v5.0.
Nick
On Tue, 2010-06-22 at 11:31 -0400, Nick Schmansky wrote:
Dear Freesurfer experts,
I have problems concerning the results of a MC simulation. When I checked the results with Qdec I discovered a matching problem for the right hemisphere. It seems like the used command runs the correct analyses for the right hemisphere, but fails projecting it properly to that one. It looks like freesurfer uses atlas and coordinates of the left hemisphere for the right hemisphere. I think the problem starts during MC simulation. Here's a "copy paste" entry of the *.log file, where I think freesurfer clearly mixes up hemisphere information:
version $Id: mri_surfcluster.c,v 1.39.2.3 2009/01/15 01:06:01 greve Exp $ hemi = lh srcid = /home/hiwibp14/smb4k/CORTICAL/fmrt/Studie43/freesurfer/qdec/Untitled/rh-Avg-Intercept-thickness/sig.mgh
For me obviously in the simulation process the programm mixes up information for right and left hemisphere during the analyses. This happens several times during analyses but only for analyses concerning the right hemisphere (see the complete attached log file). Is there any possibility to specify the command, adressing this problem like a special flag or something? Or is it possible to run the right hemisphere again separately?
To run the simulation I used the following command line:
mri_glmfit-sim \ --glmdir /home/hiwibp14/smb4k/CORTICAL/fmrt/Studie43/freesurfer/qdec/Untitled \ --sim mc-z 5000 1.3 mc-z.abs.1.3 \ --sim-sign abs \ --overwrite
Thanks for your help and best regards,
Johanna Feiler ___________________________________________________________ WEB.DE DSL ab 19,99 Euro/Monat. Bis zu 150,- Euro Startguthaben und 50,- Euro Geldprämie inklusive! https://freundschaftswerbung.web.de
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Johanna,
For the problem you describe in your email, one possibility is that the 'Untitled' directory, which is the default name qdec gives to an analysis, was overwritten with data from another analysis while the simulation script was running. The recommended procedure is to give a particular analysis its own Design name (see text box in the Design tab), using a different Design name for each hemi, and each set of factor selections.
Nick
On Tue, 2010-06-22 at 11:31 -0400, Nick Schmansky wrote:
Dear Freesurfer experts,
I have problems concerning the results of a MC simulation. When I checked the results with Qdec I discovered a matching problem for the right hemisphere. It seems like the used command runs the correct analyses for the right hemisphere, but fails projecting it properly to that one. It looks like freesurfer uses atlas and coordinates of the left hemisphere for the right hemisphere. I think the problem starts during MC simulation. Here's a "copy paste" entry of the *.log file, where I think freesurfer clearly mixes up hemisphere information:
version $Id: mri_surfcluster.c,v 1.39.2.3 2009/01/15 01:06:01 greve Exp $ hemi = lh srcid = /home/hiwibp14/smb4k/CORTICAL/fmrt/Studie43/freesurfer/qdec/Untitled/rh-Avg-Intercept-thickness/sig.mgh
For me obviously in the simulation process the programm mixes up information for right and left hemisphere during the analyses. This happens several times during analyses but only for analyses concerning the right hemisphere (see the complete attached log file). Is there any possibility to specify the command, adressing this problem like a special flag or something? Or is it possible to run the right hemisphere again separately?
To run the simulation I used the following command line:
mri_glmfit-sim \ --glmdir /home/hiwibp14/smb4k/CORTICAL/fmrt/Studie43/freesurfer/qdec/Untitled \ --sim mc-z 5000 1.3 mc-z.abs.1.3 \ --sim-sign abs \ --overwrite
Thanks for your help and best regards,
Johanna Feiler ___________________________________________________________ WEB.DE DSL ab 19,99 Euro/Monat. Bis zu 150,- Euro Startguthaben und 50,- Euro Geldprämie inklusive! https://freundschaftswerbung.web.de
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Nick Schmansky