Hello Freesurfers,
We are doing functional connectivity analyses (http://surfer.nmr.mgh.harvard.edu/fswiki/FsFastFunctionalConnectivityWalkthr...). We are on the group analysis step of step 8.
On our data set, we ran the following glmfit command:
mri_glmfit --y ./mgh/v.amy.mgh --fsgd ./FSGD/staibase28.fsgd doss --mask ./masks/gm_mask.mgz --C baseline.mtx --glmdir ./glmdir/v.staiamybaseline28.glmdir
Then we corrected using this command:
mri_glmfit-sim --glmdir ./glmdir/v.staiamybaseline28.glmdir --sim mc-z 10000 2 mc-z.absolute --sim-sign abs --cwpvalthresh 0.99 --overwrite
However, when we looked at the results, the mc-z.absolute.sig.cluster.summary, mc-z.absolute.sig.cluster.mgh, and mc-z.absolute.sig.ocn.mgh files were all different. Which of these should match the .summary file? Should they all differ? Is the summary file the final corrected results we should report? If so, what does it mean that the other files do not match?
We viewed the latter two files in tkmedit using the following command:
tkmedit fsaverage orig.mgz -aparc+aseg -overlay ./glmdir/v.staiamybaseline28.glmdir/baseline/mc-z.absolute.sig.cluster.mgh -fminmax 1.3 3
Thanks for the help!
Best, Omar
When you say they don't match, what do you mean? You have set the cwpvalthresh to .99, so when you run tkmedit, you should set the -fminmax to something like .01 3 otherwise you will not see all the clusters you see in the summary table
doug
On 3/19/14 6:28 PM, Omar Singleton wrote:
Hello Freesurfers,
We are doing functional connectivity analyses (http://surfer.nmr.mgh.harvard.edu/fswiki/FsFastFunctionalConnectivityWalkthr...). We are on the group analysis step of step 8.
On our data set, we ran the following glmfit command:
mri_glmfit --y ./mgh/v.amy.mgh --fsgd ./FSGD/staibase28.fsgd doss --mask ./masks/gm_mask.mgz --C baseline.mtx --glmdir ./glmdir/v.staiamybaseline28.glmdir
Then we corrected using this command:
mri_glmfit-sim --glmdir ./glmdir/v.staiamybaseline28.glmdir --sim mc-z 10000 2 mc-z.absolute --sim-sign abs --cwpvalthresh 0.99 --overwrite
However, when we looked at the results, the mc-z.absolute.sig.cluster.summary, mc-z.absolute.sig.cluster.mgh, and mc-z.absolute.sig.ocn.mgh files were all different. Which of these should match the .summary file? Should they all differ? Is the summary file the final corrected results we should report? If so, what does it mean that the other files do not match?
We viewed the latter two files in tkmedit using the following command:
tkmedit fsaverage orig.mgz -aparc+aseg -overlay ./glmdir/v.staiamybaseline28.glmdir/baseline/mc-z.absolute.sig.cluster.mgh -fminmax 1.3 3
Thanks for the help!
Best, Omar _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
I mean that when we look at the .summary file there are 4 clusters, only one being significant, but when we look at the mc-z.absolute.sig.cluster.mgh file there are additional clusters. The same is true for the mc-z.absolute.sig.ocn.mgh file. Shouldn't these only show the corrected clusters at the threshold of 0.05 when we view them?
You can view them here /autofs/cluster/lazar/projects/Noel/BHI/SUBJECTS_DIR/glmdir/v.staiamybaseline28.glmdir/baseline/mc-z.absolute.sig.cluster.mgh
-Omar
When you say they don't match, what do you mean? You have set the cwpvalthresh to .99, so when you run tkmedit, you should set the -fminmax to something like .01 3 otherwise you will not see all the clusters you see in the summary table
doug
On 3/19/14 6:28 PM, Omar Singleton wrote:
Hello Freesurfers,
We are doing functional connectivity analyses (http://surfer.nmr.mgh.harvard.edu/fswiki/FsFastFunctionalConnectivityWalkthr...). We are on the group analysis step of step 8.
On our data set, we ran the following glmfit command:
mri_glmfit --y ./mgh/v.amy.mgh --fsgd ./FSGD/staibase28.fsgd doss --mask ./masks/gm_mask.mgz --C baseline.mtx --glmdir ./glmdir/v.staiamybaseline28.glmdir
Then we corrected using this command:
mri_glmfit-sim --glmdir ./glmdir/v.staiamybaseline28.glmdir --sim mc-z 10000 2 mc-z.absolute --sim-sign abs --cwpvalthresh 0.99 --overwrite
However, when we looked at the results, the mc-z.absolute.sig.cluster.summary, mc-z.absolute.sig.cluster.mgh, and mc-z.absolute.sig.ocn.mgh files were all different. Which of these should match the .summary file? Should they all differ? Is the summary file the final corrected results we should report? If so, what does it mean that the other files do not match?
We viewed the latter two files in tkmedit using the following command:
tkmedit fsaverage orig.mgz -aparc+aseg -overlay ./glmdir/v.staiamybaseline28.glmdir/baseline/mc-z.absolute.sig.cluster.mgh -fminmax 1.3 3
Thanks for the help!
Best, Omar _______________________________________________ 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
The ocn will have all clusters found in the summary file. Are you seeing more than that? What you see in the sig.cluster file depends on what threshold you set when you run tkmedit. Below, you have 1.3 = p < .05, so it should only shows clusters in the summary file that have p < .05
doug
On 3/19/14 8:59 PM, Omar Singleton wrote:
I mean that when we look at the .summary file there are 4 clusters, only one being significant, but when we look at the mc-z.absolute.sig.cluster.mgh file there are additional clusters. The same is true for the mc-z.absolute.sig.ocn.mgh file. Shouldn't these only show the corrected clusters at the threshold of 0.05 when we view them?
You can view them here /autofs/cluster/lazar/projects/Noel/BHI/SUBJECTS_DIR/glmdir/v.staiamybaseline28.glmdir/baseline/mc-z.absolute.sig.cluster.mgh
-Omar
When you say they don't match, what do you mean? You have set the cwpvalthresh to .99, so when you run tkmedit, you should set the -fminmax to something like .01 3 otherwise you will not see all the clusters you see in the summary table
doug
On 3/19/14 6:28 PM, Omar Singleton wrote:
Hello Freesurfers,
We are doing functional connectivity analyses (http://surfer.nmr.mgh.harvard.edu/fswiki/FsFastFunctionalConnectivityWalkthr...). We are on the group analysis step of step 8.
On our data set, we ran the following glmfit command:
mri_glmfit --y ./mgh/v.amy.mgh --fsgd ./FSGD/staibase28.fsgd doss --mask ./masks/gm_mask.mgz --C baseline.mtx --glmdir ./glmdir/v.staiamybaseline28.glmdir
Then we corrected using this command:
mri_glmfit-sim --glmdir ./glmdir/v.staiamybaseline28.glmdir --sim mc-z 10000 2 mc-z.absolute --sim-sign abs --cwpvalthresh 0.99 --overwrite
However, when we looked at the results, the mc-z.absolute.sig.cluster.summary, mc-z.absolute.sig.cluster.mgh, and mc-z.absolute.sig.ocn.mgh files were all different. Which of these should match the .summary file? Should they all differ? Is the summary file the final corrected results we should report? If so, what does it mean that the other files do not match?
We viewed the latter two files in tkmedit using the following command:
tkmedit fsaverage orig.mgz -aparc+aseg -overlay ./glmdir/v.staiamybaseline28.glmdir/baseline/mc-z.absolute.sig.cluster.mgh -fminmax 1.3 3
Thanks for the help!
Best, Omar _______________________________________________ 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
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Douglas Greve -
Omar Singleton