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
Hi Gaurav-
Checkout the -regress_opts_3dD flag. Anything passed after that will go straight to 3dDeconvolve.
Best,
Peter
by
Peter Molfese
-
AFNI Message Board
I'm going to blame sleep depravation for the oversight on my part. Works as expected!
by
Peter Molfese
-
AFNI Message Board
Running into an odd circumstance that I can't find documentation of. I have two conditions, with times shown below on an long run (1455 TRs, TR=2 sec).
3 271 452 546 727 995 1089 1357
90 184 365 633 814 908 1176 1270
For whatever reason, 3dDeconvolve seems to not be modeling the later times after about 750ish seconds. Design Matrix link below:
Example
I've tried as
by
Peter Molfese
-
AFNI Message Board
Looks like CRAN is only supplying afex for 3.1.0 and greater. I would update R and give it another shot.
by
Peter Molfese
-
AFNI Message Board
I'll add that the T2 "structural image" (usually a 3D volume, something like a Siemens T2-SPACE) alignment to AC-PC is useful if you find your scans getting pushed to the edge of the image space. I've seen images get cutoff during TORTOISE processing, where you might lose part of the superior parts of the brain. AC-PC alignment of either the T2 image or the template image wi
by
Peter Molfese
-
AFNI Message Board
The most straightforward approach is to use DIFF_CALC (calcvm) to open the *_DMC.list for each processed DWI file. From there you can export the data to AFNI, which will create NIFTI format files (DWI.nii). You can then use 3dDWItoDT to fit the tensors and continue through FATCAT. You can optionally fit the tensors in TORTOISE using a variety of algorithms (linear, nonlinear, RESTORE, iRESTORE
by
Peter Molfese
-
AFNI Message Board
That error is usually caused by Freesurfer not being loaded on the path. The typical way one does that with Freesurfer is something like this:
Bash:
export FREESURFER_HOME=/Applications/mri/freesurfer
source $FREESURFER_HOME/SetUpFreeSurfer.sh
tcsh:
setenv FREESURFER_HOME /usr/local/freesurfer
source $FREESURFER_HOME/SetUpFreeSurfer.csh
by
Peter Molfese
-
AFNI Message Board
Hi Ajay,
Not sure about the best way to go about flipping a cortical surface. Since the process of converting surfaces to GIFTI and resampling to the fsaverage are all based on Freesurfer programs, I would suggest asking the folks over at MGH/Martinos on the preferred way to do that. The HCP group might also have thoughts on ways of handling this.
Once you get that L/R flip, then as lon
by
Peter Molfese
-
AFNI Message Board
For the sake of it being Friday afternoon, it being easier, and also the method that I use in my lab, I would recommend option 1.
The output of the mris_convert command is transformed to the fsaverage space. So there is no need to create standard meshes. Standard meshes created using MapIcosahedron are another way to get all of the data ready for group analysis by getting vertex correspond
by
Peter Molfese
-
AFNI Message Board
I generally recommend the step() function. But to continue using the "iszero":
3dcalc -a file.nii -expr 'or(iszero(a-3), iszero(a-6), iszero(a-25))' -prefix maskimage
by
Peter Molfese
-
AFNI Message Board
Hi Ajay-
As you've discovered, @SUMA_Make_Spec_FS does not convert the outputs of -qcache. There are two somewhat direct ways of doing cortical thickness analysis in AFNI/SUMA. The first is to use the -qcache outputs and convert using the mris_convert command you listed for each subject.
mris_convert -c ./Subject01/surf/lh.thickness.fwhm10.fsaverage.mgh \
$SUBJECTS_DIR/fsaverage
by
Peter Molfese
-
AFNI Message Board
By default AFNI uses RAI coordinates. When you click on the image and say "Jump to XYZ", you may notice that it says "RAI" in the top of that window. You can change to "SPM" or LPI coordinates by right clicking on the upper left side of the controller window (where it says "order: RAI=DICOM") and selecting LPI. Then your jump to command will say "LP
by
Peter Molfese
-
AFNI Message Board
Sorry, I meant to get back to you on this earlier! Can you provide a sample set of commands?
My go-to solution for ROIs (because we run it anyway, and have gotten very good at it) is to use Freesurfer to do a segmentation and use the individual segmentations as our ROIs.
by
Peter Molfese
-
AFNI Message Board
Two things to check:
1) have you tried just using 3dNwarpApply to skip the NwarpCat step? You can concatenate warps and invert them within that program.
2) On images that end up cropped, is @auto_tlrc reporting a larger CENTER_DISTANCE than on images that are not being cropped?
by
Peter Molfese
-
AFNI Message Board
I suspect Paul will respond soon. But when I've run analyses that don't show any significant tracts, it's usually because the ROI masks (netrois) are outside of the brain area. How does the overlap of the ROIs you want tracked look overlaid onto of the 3dDWItoDT output?
How did you make your ROIs? Using an Atlas? Overlaid on an anatomical?
Can you provide your exact com
by
Peter Molfese
-
AFNI Message Board
Can you post the exact commands you used and the output of afni -ver?
by
Peter Molfese
-
AFNI Message Board
Hi Floris-
When using the -nodata flag, you can supply the onset times using the standard -stim_times format. A recently used setup of my own went something like this:
3dDeconvolve -nodata 290 2 -polort 3 -num_stimts 1 -stim_times 1 condA.txt 'GAM' -x1D X.xmat.1D -xjpeg X.jpg
Those times can be in seconds however you would like to specify them (not just TR markers). It is, h
by
Peter Molfese
-
AFNI Message Board
Sorry, I was apparently distracted mid-repose! Gang's approach would work.
by
Peter Molfese
-
AFNI Message Board
I suggest you use option #2. You can then use 3dttest++ with the -paired option.
by
Peter Molfese
-
AFNI Message Board
Your version of AFNI is fairly old. It would be worth updating (@update.afni.binaries -d) to see if that fixes the problem. Also you can add -volreg_align_e2a to make the script setup the alignment step as EPI to Anatomic.
Note that if you are using the NeuroDebian binaries, you will need to uninstall afni using apt-get and then download the binaries "openmp" binaries from this s
by
Peter Molfese
-
AFNI Message Board
Can you give us the output of:
afni_system_check.py -check_all
Did you follow these instructions? Andy put them into video format too.
by
Peter Molfese
-
AFNI Message Board
I also receive the error. But adding a mask (-mask AUTO) fixes it.
by
Peter Molfese
-
AFNI Message Board
ConvertDset should work on a GIFTI file. You'll need to pad the dataset using -pad_to_node. See these two threads:
1. Useful Syntax -
2. What value to place in -pad_to_node -
by
Peter Molfese
-
AFNI Message Board
Try something along these lines:
SurfClust \
-i fsaverage/SUMA/lh.pial.asc \
-input 'lh.SD_vs_HV_ageGenderCov.gii[1]' 0 \
-athresh 2.66 \
-rmm -1 \
-out_roidset \
-prefix test
by
Peter Molfese
-
AFNI Message Board
You could checkout SurfFWHM. It has given me believable values for fMRI results.
by
Peter Molfese
-
AFNI Message Board
Hi Giovanni,
My intuition is to use the fsaverage_SurfVol+orig. as your surf_vol file.
As for the level of blur, did you use the blurred files in Freesurfer (like I did here) or did you use SurfSmooth?
I believe that the Freesurfer method in mris_preproc adds smoothness, instead of the "smooth to a level of" like 3dBlurToFWHM does. In the Freesurfer case, you would have to
by
Peter Molfese
-
AFNI Message Board
The program you are looking for is slow_surf_clustsim.py
by
Peter Molfese
-
AFNI Message Board
As Rick said, this isn't yet implemented. In my use of stim_times_IM, we've resorted to getting the sub-brick labels:
3dAttribute -ssep ' ' -name BRICK_LABS dset+tlrc
And then divided it up using Python. Some relevant functions: string.split(), string.sort(), ",".join(string). Take the output string for running your 3dmaskave command.
Truthfully, the m
by
Peter Molfese
-
AFNI Message Board
The $ is effectively a "the last sub-brick" selector. So you could try:
stats.subj1_REML+tlrc'[1..$(2)]'
More info is available near the end of the afni -help output.
INPUT DATASET NAMES
-------------------
An input dataset is specified using one of these forms:
'prefix+view', 'prefix+view.HEAD', or 'prefix+view.BRIK'.
You
by
Peter Molfese
-
AFNI Message Board