Show all posts by user
Dear AFNI users-
We are very pleased to announce that the new AFNI Message Board framework is up! Please join us at:
https://discuss.afni.nimh.nih.gov
Existing user accounts have been migrated, so returning users can login by requesting a password reset. New users can create accounts, as well, through a standard account creation process. Please note that these setup emails might initially go to spam folders (esp. for NIH users!), so please check those locations in the beginning.
The current Message Board discussion threads have been migrated to the new framework. The current Message Board will remain visible, but read-only, for a little while.
Sincerely,
AFNI HQ
History of AFNI updates
Results 4501 - 4524 of 4524
Hi Georgios,
My suggestion is what afni_proc.py does, of course. It truncates the
minimum EPI dimension to 3 significant bits, and uses isotropic voxels.
In your case, it would use 3 mm voxels, basically matching the original
voxel dimensions. Since rotations can rotate the 3 mm direction into
the 6 mm axis, I like throwing away the big dimension.
See the afni_proc.py help output for
by
rick reynolds
-
AFNI Message Board
Hi Han,
Yes, I meant to leave the afni_proc.py processing as is, and to
run the command that I gave you basically as is. The anat_final
dataset is from the afni_proc.py script. Just apply that to the
original anat.
- rick
by
rick reynolds
-
AFNI Message Board
So this is from an afni_proc.py script?
In any case, you can apply @auto_tlrc with usage 2 to transform the
original anatomy by the matrix in the +tlrc result. For example:
@auto_tlrc -apar anat_final.SUBJ+tlrc -input SUBJ_anat+orig
Here anat_final+tlrc is the skull-stripped version in standard space, and
SUBJ_anat is the initial version with the skull. The result should be
SUBJ_ana
by
rick reynolds
-
AFNI Message Board
Hi Nic,
Having different orientations is not a big problem. As you mentioned,
3dresample can easily fix that (without any actual interpolation, since
the voxel coordinates don't change). To do that, I would use -orient
instead of -master (though they ought to imply the same thing in such a
case).
However, since with -master one of the images moves out of view, that
suggests the
by
rick reynolds
-
AFNI Message Board
Hi Andrew,
1. Yes. The dmBLOCK(0) default makes sense, as we generally
expect longer stimulus responses to behave as a convolution
of a short response and a top hat function. That is actually
a somewhat recent change, see the output from:
afni_history -past_months 2 -author rwcox -program 3dDeconvolve
2. That is why dmBLOCK(1) is not the default, because dmBLOCK(0)
make more sens
by
rick reynolds
-
AFNI Message Board
Note the help. The usage is: 3dUndump infile(s).
All options must come before the input file(s). In your case,
roi_PCC.txt is the input file. So -srad must come first.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Anjali,
At least the second error message makes it clear that you should
not transpose the ideal file.
Regarding the first error:
1. Is that -input really the correct file name? It almost looks
like 2 names together. Maybe.
2. Try running dos2unix on the files. Maybe they have non-unix
characters in them. You can use file_tool to check that.
I see your AFNI version
by
rick reynolds
-
AFNI Message Board
I would not call y the observed BOLD response, but rather the observed
MRI signal. It is a combination of many things, which one tries to model
as well as possible (using linear regression with 3dDeconvolve, for
example). Often it might be expressed as some linear sum of baseline,
motion, responses to known stimuli, and remaining noise.
The 'y' in that equation is the EPI time
by
rick reynolds
-
AFNI Message Board
Hi Phil,
I'm happy to hear that afni_proc.py is working well for you.
The iresp files should be normally viewable in suma. While
the view on the surface might not be so interesting, you can
keep a graph window open to view the iresp time series for
the current node.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Nobuko,
Note that every subject has the same mesh, and that the surfaces
are considered to be in alignment (though the shapes vary, nodes
are supposed to be in basically the same place on the cortex).
Anyway, yes, pick a representative subject.
---
I am referring to -amm2 in SurfClust.
---
The slow_surf_clustsim.py program is to help determine a threshold
and cluster size
by
rick reynolds
-
AFNI Message Board
Hi John,
I suspect you did not use a mask. Note that you cannot use -automask
on the errts time series (since the means are 0, it won't find one).
Try applying a mask generated from the volreg output.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Gregory,
This looks like you are trying to do a -1/1 amplitude modulation.
Maybe it would be enough to "marry" your stimulus timing files
to the -1/1 values, and run 3dDeconvolve using -stim_times_AM1
for that regressor class.
You would not need timing_tool.py or waver.
- rick
by
rick reynolds
-
AFNI Message Board
That package is more than 4 years old. A virtual system might
be a far better way to go.
But if you really want it, the cygwin package is under here:
http://afni.nimh.nih.gov/pub/dist/tgz/old_binaires
- rick
by
rick reynolds
-
AFNI Message Board
Hi Ece,
I did make a small change to allow for this, however I think this is not a great place to
do it. In your case, I think you should catenate the 1D files and use -concat.
However (once the update is distributed), your command should work, with a couple of
changes: 1. since you are abusing the 3D functions handling 1D files, your datasets
need to be transposed on input, and 2. -for
by
rick reynolds
-
AFNI Message Board
Hi Ece,
I think auto-catenation is used for multiple datasets, but maybe that does
not work on 1D input.
While we ponder this, you might use 3dTcat to catenate those input datasets,
then add the -concat option to specify run breaks.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Sandra,
Voting does not seem like a good idea, since voxels will not have the same values
(since it's a probability mask), unless there is a region where they are all 1. And your
clip at 0.8 for the gray matter makes it even less likely to work.
What you might do is either use linear (or whatever) interpolation in forming the new
grid, and then threshold to form a binary mask, o
by
rick reynolds
-
AFNI Message Board
Hi Ben,
If you are using the Clusterize GUI, then apply the SaveMsk option.
That would write a dataset where each cluster gets its own value
(from 1=biggest to N=smallest). To make a dataset of just one cluster,
use 3dcalc to make a new dataset with just that value.
Let me know if you need more details.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Yiyu,
This seems a little odd. Would you please mail me directly?
- rick
by
rick reynolds
-
AFNI Message Board
Note that the beta estimates are (generally) per regressor, not per time interval.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Krishna,
I did get around to making changes for this (more than a week ago, actually).
So when we next run a build (hopefully tonight), they should be available.
- rick
by
rick reynolds
-
AFNI Message Board
Hi David,
I see what you mean, but do not trust the alignment shown by such images
(where one cannot follow the cortical folds of the datasets). Actually, the way
your older images follow the contours worries me (perhaps you used mutual
information for the cost function?).
My expectation is that the new alignment is good. If you would like me to take
a closer look at a couple of sub
by
rick reynolds
-
AFNI Message Board
Hi Danielle,
Are you working with Yiyu?
What version of python are you running? It is possible
that the form of 'except' that is on the previous line
has been depreciated.
- rick
by
rick reynolds
-
AFNI Message Board
Hi YIYU,
Hmmm, maybe that syntax was depreciated. What version
of python are you running (python -V)?
- rick
by
rick reynolds
-
AFNI Message Board
Hi Tim,
Consider using afni_proc.py to do the censoring, and then
see how it does it in the processing script. By default,
afni_proc.py suggests censoring at a masked fraction of 0.1.
See "afni_proc.py -help" for details, and consider an option
like '-regress_censor_outliers 0.1', and possibly along with
an option like '-regress_censor_motion 0.3'.
Afte
by
rick reynolds
-
AFNI Message Board