Hi Douglas,
I used the matlab code and the p value is signifiant, although less so than the
p-value I get from the clusterwise statistics (0.0027 vs 0.00010). (Indeed I
used an abs threshold for the clusterwise statistics).
Which p-value should I report?
Thanks,
Maaike
From: r_maa...@hotmail
Hi,
I am new to FreeSurfer. I am trying to run recon-all using directives
-autorecon1 and -autorecon2 through a python script using a module called
subprocess that allows me to call command line commands.
Unfortunately, I am getting the following error:
*mghRead(/home/roberto/Documents/FreeSurfe
Hi Yawu
the recon-all.log indicates that building the base finished without
error. What command did you run that gave you the error? The original
nifti files will have different geometries since the subject will be in a
different location in the scanner, but that should be normalized out by th
Hi Bruce,
I run recon-all -base K_El03_base -tp K_EL03_1 -tp K_EL03_2 -all.
It gives a warning, you can find it from the recon-all.log :
mri_diff --notallow-pix
/usr/local/src/freesurfer/subjects/K_EL03_2/mri/rawavg.mgz
/usr/local/src/freesurfer/subjects/K_EL03_1/mri/rawavg.mgz
Volumes differ
Hi Yawu,
the base stream tests if the header geometries are the same across time points.
For example, it is a bad idea to change the resolution (voxel size) of your
inputs across time. Can you send the output of
mri_info /usr/local/src/freesurfer/subjects/K_EL03_1/mri/rawavg.mgz
and
mri_in
Hi Yawu,
I looked it up based on the data you send us. You are fine, your voxel sizes
and image dimensions are the same, only the RAS coords differ and that is OK.
I'll see if I can update the test in a way that it does not warn in these cases.
Best, Martin
On Jan 30, 2015, at 9:47 AM, Martin
Dear FS Experts,
In order to generate the statistics using mri_segstats I ran the following
command line:
mri_segstats --seg${in}/${i}/mri/aparc+aseg.mgz --ctab-default --i
${insa}/${i}/fa.anat.nii --mask ${insa}/${i}//svs.anat.nii --sum
${insa}/${i}/fa.summary.dat
In the final report (fa.sum
I have no idea what the problem is. You can have it exclude the fsnr
maps by adding -no-anamaps, though there might be something else going
on with the data. Can you give me the location of your data and
SUBJECTS_DIR as well as the command line you ran and the full terminal
output of the comma
You should report the clusterwise p-value. The p-value that you have
computed is not interpretable. If it did not come out significant, then
it would be worrisome, but it is only a check and cannot be used for
anything. What p-value did you get from SPSS? I'd be curious to know
what the sourc
There is not a flag to control the printing format. However, you can
multiply the values by a constant (--mul flag). Eg, you can multiply all
values by 100 or divide by 100.
doug
On 1/30/15 11:42 AM, Alshikho, Mohamad J. wrote:
> Dear FS Experts,
> In order to generate the statistics using mri_
The p value from SPSS is 0.156. I checked with a statistician, just to be sure,
but he says the SPSS model is correct... (he doesn't know anything about
freesurfer and FSGD files, but I told him it the model is a GLM with age,
gender and state as covariates).
Date: Fri, 30 Jan 2015 11:46:37 -050
Dear all,
We are trying to make a new contrast for our GLM analysis, however we have the
following flag error.
Is there any solution for this?
Thank you
isik
[schnook:subjects] (nmr-stable53-env) mkcontrast-sess -an
RM_hard_easy_fake_error -co hard_v_fake+easy -a 1 -c 2 -c 3
(standard_in) 1
Can you run the following cmd and send me the terminal output?
mkcontrast -debug -anadir RM_hard_easy_fake_error -wcond 1.
-.5000 -.5000 0 -sumconds -o
RM_hard_easy_fake_error/hard_v_fake+easy.mat -wdelay 1 -no-octave
On 1/30/15 12:15 PM, Kar
>>>I told him it the model is a GLM with age, gender and state as
covariates
This is not a sufficient description. Several design matrices could be
derived from this description.
On 1/30/15 11:58 AM, maaike rive wrote:
The p value from SPSS is 0.156. I checked with a statistician, just to
be
How should I have described the model?
Date: Fri, 30 Jan 2015 13:09:40 -0500
From: gr...@nmr.mgh.harvard.edu
To: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] average values per cluster
>>>I told him it the model is a GLM with age, gender and
state as covariates
Here it is:
[schnook:subjects] (nmr-stable53-env) mkcontrast -debug -anadir
RM_hard_easy_fake_error -wcond 1. -.5000
-.5000 0 -sumconds -o
RM_hard_easy_fake_error/hard_v_fake+easy.mat -wdelay 1 -no-octave
$Id: mkcontrast,v 1.10.2.1 2011/0
It depends on the actual covariates and whether you used DOSS or DODS.
The default is DODS, which is a full interaction model. Assuming that
state is a categorical variable with 2 levels, you'd have 4 regressors
(gender by state) + 4 more regressors (gender by state by age). If the
SPSS model
sorry, can you send me your mkanalysis-sess command line as well? You
can find it in the analysis.info file in the analysis folder
On 1/30/15 12:15 PM, Karahanoglu, Fikret Isik wrote:
>
> Dear all,
>
> We are trying to make a new contrast for our GLM analysis, however we have
> the following fl
Hi all,
I've seen very similar questions posted, but I haven't seen one with a
response.
I just installed freesurfer onto a mac os x. I did the following steps:
export FREESURFER_HOME=/Applications/freesurfer
source $FREESURFER_HOME/SetUpFreeSurfer.sh
and got this output:
freesurfer-D
OK, we will try with -no-anamaps flag.
the subjects_dir: /cluster/manoach/AS/RespMon12C/subjects
I attach our simple code and the log output.
Thank you,
bengi and isik
From: freesurfer-boun...@nmr.mgh.harvard.edu
[freesurfer-boun...@nmr.mgh.harvard.ed
The sudo is causing it to run under a different user (root) which means
that it opens a new shell that does not have the proper environment set.
Why are you running it as root? If you set SUBJECTS_DIR to something
other than
/Applications/freesurfer/subjects you won't have to run as root
d
So we ran
mkcontrast-sess -an RM_hard_easy_fake_error -co hard_v_fake+easy -a 1 -c 2 -c 3
from the terminal. I attach the mkcontrast-sess.log and anlaysis.info.
isik
From: freesurfer-boun...@nmr.mgh.harvard.edu
[freesurfer-boun...@nmr.mgh.harvard.edu] o
sorry, I wanted the mkanalysis-sess cmd, not the mkcontrast-sess
On 1/30/15 1:50 PM, Karahanoglu, Fikret Isik wrote:
So we ran
mkcontrast-sess -an RM_hard_easy_fake_error -co hard_v_fake+easy -a 1 -c 2 -c 3
from the terminal. I attach the mkcontrast-sess.log and anlaysis.info.
isik
___
There is an error msg in the log file:
cat:
/autofs/cluster/manoach/AS/RespMon12C/subjects/RMSCH012/subjectname: No
such file or directory
Can you check whether this file exists? I don't have permission to view it.
doug
On 1/30/15 1:43 PM, Karahanoglu, Fikret Isik wrote:
OK, we will try
Ok, I'm so sorry, I'm afraid I mixed things up. I indeed used DODS, and assumed
these were may regressors:
diagn1*state1*gender1 diagn1*state1*gender2 diagn1*state2*gender1
diagn1*state2*gender2 diagn2*state1*gender1 diagn2*state1*gender2
diagn2*state2*gender1 diagn2*state2*gen
Actually, that looks correct. Is this what you fed to SPSS?
On 1/30/15 2:07 PM, maaike rive wrote:
Ok, I'm so sorry, I'm afraid I mixed things up. I indeed used DODS,
and assumed these were may regressors:
diagn1*state1*gender1
diagn1*state1*gender2
diagn1*state2*gender1
diagn1*state2*gend
I thought so, but I will check with the statistician next Monday, I'll get back
to you as soon as I'm sure.
Date: Fri, 30 Jan 2015 14:12:32 -0500
From: gr...@nmr.mgh.harvard.edu
To: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] average values per cluster
Actu
We are just adding new analyses to a previous study, apparently, they didn't
run mkanalysis-sess. But, instead, they modified
analysis.info and analysis.cfg (attached) from another study with the same
design. But we had no problem before with other contrasts...
isik
This is analysis.info comes from an old version of FS and will not work
with 5.3. When you made new contrasts before, are you sure you used 5.3?
doug
On 1/30/15 2:26 PM, Karahanoglu, Fikret Isik wrote:
We are just adding new analyses to a previous study, apparently, they didn't
run mkanalysi
We ran individual level analysis with the existing contrasts, which didn't
yield any incompatibility issues.
So all other contrasts must have been created with old version.
At that point what do you recommend? We continue with old version or redo
everything with new version?
isik
__
On second thought, I think the reason for the discrepancy is that I included
state as a factor instead of covariate in the SPSS model. I ran the model again
after having adjusted this and included every possible interaction; now I reach
a p value of 0.009 (which is still different but at least s
Thanks so much for noticing that. We added the subjectname file and ran again
with -no-anamaps flag.
It finished without any errors though we only get ces. outputs, not the tal.
outputs. Is it due to the flag?
Again here is the log file.
isik
From: frees
Thanks so much for the fast reply Doug!
I changed the directory for freesurfer home...so now if I source, it
returns:
source $FREESURFER_HOME/SetUpFreeSurfer.sh
freesurfer-Darwin-lion-stable-pub-v5.3.0
Setting up environment for FreeSurfer/FS-FAST (and FSL)
FREESURFER_HOME /Us
Do you have write perms to this file and the folder it is in?
/Users/Tyler/Documents/freesurfer/subjects/bert/scripts/recon-all.env to
/Users/Tyler/Documents/freesurfer/subjects/bert/scripts/recon-all.env.bak
On 1/30/15 3:09 PM, Tyler Good wrote:
Thanks so much for the fast reply Doug!
I cha
I would redo everything with the new version. If you're mixing versions,
I can't give you recommendations on what would work or not.
On 1/30/15 2:53 PM, Karahanoglu, Fikret Isik wrote:
> We ran individual level analysis with the existing contrasts, which didn't
> yield any incompatibility issue
Hi Doug,
It looks I have rw perms for the both of those files and the folders they
are in.
?
Thanks!
On Fri, Jan 30, 2015 at 3:11 PM, Douglas Greve
wrote:
>
> Do you have write perms to this file and the folder it is in?
> /Users/Tyler/Documents/freesurfer/subjects/bert/scripts/recon-all.env to
Some where in there you don't have perms, it might be one of the higher
folders, try
chmod -R u+rwX /Users/Tyler/Documents/freesurfer/subjects
On 1/30/15 3:50 PM, Tyler Good wrote:
Hi Doug,
It looks I have rw perms for the both of those files and the folders
they are in.
?
Thanks!
On F
Hello, first time posting after several google searches. Could anyone help to
answer if there is a way to have a detailed reconstruction of the insula cortex
with the gyral and sulcal anatomy? Details would include thickness of each
sulcus and gyrus etc?
How could this be achieved if there is s
38 matches
Mail list logo
