Hi,
We see that particular WARNING message all the time and simply ignore it.
It is simply telling you that var_filtered_func_data has some zeros (or
negative) values where mask.nii.gz is non-zero. If you look at where
these voxels are, you should find that they are just a rim around the
brain. This arises (I think) because of some peculiarities (which you
could identify by working through the steps in the logs) regarding how the
mask gets defined in Level1 vs. higher levels.
If you run:
fslmaths var_filtered_func_data -bin -Tmin tmp
fslstats tmp -V
you should get the "nmaskvoxels" number in the report_log.html (922351 in
your example).
Note that this number will be a little bit smaller than you get from
fslstats mask -V
cheers,
-MH
--
Michael Harms, Ph.D.
-----------------------------------------------------------
Conte Center for the Neuroscience of Mental Disorders
Washington University School of Medicine
Department of Psychiatry, Box 8134
660 South Euclid Ave. Tel: 314-747-6173
St. Louis, MO 63110 Email: [log in to unmask]
On 5/30/13 10:20 AM, "Marcelo Mattar" <[log in to unmask]> wrote:
>Thanks, Matt!
>
>I just uploaded the tar'd directory on the link you sent me. A couple of
>notes to make your life easier:
>
>1) The warning message can be seen on ./copeX.feat/report_log.html, and
>it happens for all the contrasts.
>2) I had wondered if the warning was being caused due to having more EVs
>than lower level FEAT directories. It turns out that the warning is still
>there even if I run the lower and higher level models with a single EV
>and COPE. Indeed, if we look at the warning message carefully, it
>actually says that the number of evs is 1 while the number of points is
>4. An example of the warning is pasted below:
>
>WARNING: The passed in varcope file, var_filtered_func_data, contains
>voxels inside the mask with zero (or negative) values. These voxels will
>be excluded from the analysis.
>nevs=1
>ntpts=4
>ngs=1
>nvoxels=922351
>Running:
>nmaskvoxels=922351
>
>Thank you very much for your willingness to help again. Hopefully we'll
>be able to figure this out.
>
>Best,
>
>--Marcelo
________________________________
The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.
|