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 4141 - 4170 of 4524
Hi Jay,
For tcsh, replace steps 4 through 5 with:
nano .cshrc
add the text:
set path = ( $path ~/abin )
if ( $?DYLD_FALLBACK_LIBRARY_PATH ) then
setenv DYLD_FALLBACK_LIBRARY_PATH $DYLD_FALLBACK_LIBRARY_PATH:~/abin
else
setenv DYLD_FALLBACK_LIBRARY_PATH ~/abin
endif
ctrl-x and yes to save
- rick
by
rick reynolds
-
AFNI Message Board
Hi Peter,
There had been a path in the motion file up until about half a year ago.
How old are your binaries?
- rick
by
rick reynolds
-
AFNI Message Board
Hi Maria,
The trouble with a 7 second TR, starting around maybe 4 seconds,
is that the sampling is not frequent enough to capture the shape
of a BOLD curve. At 7 seconds, such a curve for one event might
have just one non-zero point. Interpolation of such a curve
would be very noisy. And since you acquire all of your data in
the first 2.5 seconds of the TR, it might be the most accura
by
rick reynolds
-
AFNI Message Board
Hi Maira,
If the experiment had no rest condition, then the reviewer would
have a point. If the only other condition besides the 6 main
working memory tasks was scrambled images, then scrambled images
would be the baseline/rest condition, so it would not be modeled.
But in your case, it is not a baseline condition. You have a
separate fixation condition for baseline.
I expect that
by
rick reynolds
-
AFNI Message Board
Hi Sharlene,
Sure, NIfTI files should work fine. The InstaCorr plugin
and 3dGroupInCorr program have no problem with such
datasets.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Wolfram,
Yes, you should be good on that front.
I hope your results come out.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Peter,
This is probably the link that I included. The second part of
the response describes why this is generally not a worry.
afni.nimh.nih.gov
- rick
by
rick reynolds
-
AFNI Message Board
Hi David,
Since that dataset seems to exist (referenced as the
'final stats dset' a few lines above), it suggests that
there is just no Full_Fstat sub-brick in the dataset,
which is strange.
Did you alter the afni_proc.py script at all, or run it
straight from the command?
If you would like, send me private email with the full
output from the proc script. It might be help
by
rick reynolds
-
AFNI Message Board
Yes, that should be all you need. And hopefully all of the
ANALYZE datasets have the same orientation, so you should
really only need to figure it out once, and then apply it to all
datasets using 3drefit.
However, I suggest you start by using 3dcopy to copy each
ANALYZE dataset into AFNI format before working on it. Not
only do we not write the ANALYZE format (no one should),
but you
by
rick reynolds
-
AFNI Message Board
If you know the anats have the E capsule on the right side, then
you can easily figure out their orientations.
Can you see the capsule in the EPI images as well?
If you would like, upload an anat and EPI dataset for some subject,
along with a readme file describing where the vitamin E capsule is,
and I will take a look.
- rick
by
rick reynolds
-
AFNI Message Board
Hi David,
A vitamin E capsule would be great, assuming you know which
side it is on. But that does not help for the old data.
The processing is not overly affected by this, but if you
cannot really tell left from right, it does not seem safe to
use the data.
Is it just the EPI that you are unsure about, the anatomy,
or both?
- rick
by
rick reynolds
-
AFNI Message Board
Hi Peter,
Sorry, afni_proc.py does not have the ability to take a mask as input.
You can insert -blur_in_mask yes anywhere.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Krishna,
To bet basically any 3dDeconvolve response model, run 3dD
with a single stimulus at time 0, -polort -1 and grab the X-matrix.
e.g.
3dDeconvolve -nodata 100 1 -polort -1 -num_stimts 1 -stim_times 1 '1D: 0' 'MION(0)' -x1D X.xmat.1D -x1D_stop
1dplot X.xmat.1D
- rick
by
rick reynolds
-
AFNI Message Board
I think the only description of AFNI_CONVMODEL_REF is in the help
for that function, which you have probably noticed in looking at the
code. But just for reference, you can get at the function help like this:
setenv AFNI_MODEL_HELP_CONVDIFFGAM Yes
3dNLfim -signal ConvDiffGam
In any case, yes, you should be able to use 1 0 0 0 0 0. But perhaps
the length of that time series should matc
by
rick reynolds
-
AFNI Message Board
Hi Wolfram,
After taking a look at your data, the problem is subtle, unfortunately.
The effect of such a large blur is that the range of values becomes
much smaller at any given voxel (i.e. the min and max of each time
series is much closer together).
That is not itself a big deal. But things go bad because the data is
stored as shorts. So everything is truncated to integers, and th
by
rick reynolds
-
AFNI Message Board
Hi Josh,
This is almost entirely due to not having a tight mask.
There are lots of values close to zero included in the
computation (maybe from masking before a blur), which
really drag down the results.
Note that when afni_proc.py does this, it restricts the
computation to the EPI mask (full_mask).
- rick
by
rick reynolds
-
AFNI Message Board
Hi Shrant,
For completeness, let me also mention that using 3dDeconvolve
for this is easy too, except that it requires one more step.
Use -polort -1 (presumably trends are already gone), provide
the one -stim_file, and use -rout. But -rout actually gives
R^2, so you will have to multiply by the sign of the beta or
t-stat. That can be added to the subsequent 3dcalc command,
as given by
by
rick reynolds
-
AFNI Message Board
Hi pt,
First of all, what's with you and Daniel having a chat in the
middle of the night?
I think Daniel is suggesting that the overlay sub-bricks become
slaves to the underlay sub-brick. So do not try to control them
by choosing OLay or Thr indices, control them via underlay
sub-brick indicies (which even applies when there is just one
underlay volume).
If there is just one u
by
rick reynolds
-
AFNI Message Board
Hi Laurel,
Your command looks okay, except that the mask is 1 mm. Are you
oversampling the data?
I will have to play around with some of this myself, and try to
reproduce these findings...
- rick
by
rick reynolds
-
AFNI Message Board
Hi Jeff,
Along the lines of what Daniel is suggesting, to do what you
suggested (scaling all mean/stdev values to be the same across
subjects), the following sort of script could be done.
1. Make the subject masks with 3dSkullStrip (or even 3dSeg, to
be fancy). If there are holes in the masks, consider using
3dmask_tool to fill them:
3dmask_tool -input $subj.anat.stripp
by
rick reynolds
-
AFNI Message Board
Maybe model_conv_diffgamma.c is close to what you want?
I see that function does not appear to be included in the output
of "3dNLfim -help", I will add it.
- rick
by
rick reynolds
-
AFNI Message Board
Hi Nic,
Since the 3dMEMA values are bounded by 100, that means you
basically get 2 decimal places in the result. So any values less
than 0.01 (well, maybe 0.003) in abs value will be truncated to 0.
But just so we understand, exactly how is this truncation a problem?
What problem is caused by values of exactly 0 in the result? Is it
that they get no color in the GUI?
- rick
by
rick reynolds
-
AFNI Message Board
Ah, I think I see the problem. Maybe.
You are passing "-polort 2" via -regress_opts_3dD, which means
afni_proc.py does not know about it (those options are directly
passed to 3dDeconvolve). So maybe AP has assumed a higher
degree (which it passed to 3dD, but yours overrode it, being later),
and applied it when selecting columns of the design matrix for ideal
files.
Instead,
by
rick reynolds
-
AFNI Message Board
Hi Liana,
As Colm notes, you may be censoring a LOT. The error
states that there are only 6 time points in the input
(after censoring). Plot the enorm file and see what
you think. Maybe this subject needs to be dropped?
- rick
by
rick reynolds
-
AFNI Message Board
Hi Liana,
I am not sure from this. Would you please send me email with
the proc script and the output text file (output text from running
the proc script)? I will send you a PM with me address.
- rick
by
rick reynolds
-
AFNI Message Board
Hi David,
Sorry for being slow on the response. Go ahead and bump your
own message if we seem to miss it.
The reason that this method might not work is that it seems
the ANALYZE datasets are not in the same orientation as the
DICOM images (e.g. I and S are reversed). That means that
the ANALYZE images have actually been transformed (or they
were not all scanned the same way).
So I
by
rick reynolds
-
AFNI Message Board
Hi Anita,
The "Voxels edited out" count is the total number of voxels
(after masking) that are removed from the result due to a
small cluster size.
I think the problem you are seeing is that the TT_N27 dataset
probably has holes in it, which are disrupting your result. Fill
in that dataset to form a more proper mask:
3dmask_tool -inputs ~/abin/TT_N27+tlrc -dilate_input 5 -
by
rick reynolds
-
AFNI Message Board
Hi Wolfram,
A) This is very strange, and is possibly related to removing
so much from the data (i.e. bandpassing). I have not
actually seen that before. What is your 3dFWHMx command?
B) The FWHM should not be less. It would be good to figure
out what is going on with that. If you cannot see what
is going wrong, maybe you could upload some data so that
I could reprodu
by
rick reynolds
-
AFNI Message Board
Hi Sam,
No one is going to give an exact answer for this, since it is
a matter of looping over your ROIs and subjects, where the
dataset names are known. But of course, we don't know what
directory structure or dataset names you are using.
In any case, this just amounts to some sort of looping:
for each subject
for each ROI
3dmaskave -quiet -mask ROI_DATASET SUBJEC
by
rick reynolds
-
AFNI Message Board