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
Page 1 of 1 Pages: 1
Results 1 - 27 of 27
Thank you, Rick. It should't take me too long to add this to the realtime plugin.
Jonathan
by
jlisinski
-
AFNI Message Board
Thanks for your reply, Gang! But what if the data is written by plug_realtime.so. You can't specify an output filename ending in .nii there. I was hoping to find an environment variable similar to AFNI_AUTOGZIP that can be set to write all data as .nii without explicitly specifying it in the filename. It's okay if it doesn't exist. I just want to make sure I didn't miss it.
by
jlisinski
-
AFNI Message Board
Hi all,
I was wondering if there is a way to default to (environment variable) writing .nii instead of HEAD/BRIK. I looked through the environment variables in README.environment, but did not find one for this purpose. Did I miss it?
Thanks for the help!
Jonathan
by
jlisinski
-
AFNI Message Board
Hi Paul,
I would use each volume (instead of betas from individual blocks) and exclude block transitions (4-6 seconds) if you have a (true) blocked design and if you expect a steady-state signal after 2-6 seconds for each block. I would only recommend 3dDeconvolve, 3dLSS, 3dTfitter, etc. as an initial step for event-related or mixed designs. We've used 3dLSS with some success for event-re
by
jlisinski
-
AFNI Message Board
Hi Paul,
The dataset did not go through any other processing to my knowledge. I agree, It should have the to3d command in the history. The fact that this it is not there is odd. I will do some more digging...
Thank you for all your help!
Jonathan
by
jlisinski
-
AFNI Message Board
Thanks - I've uploaded the .HEAD and the .BRIK.gz file. We used 18.2.10 for this, but are running 19.2.01 now (linux_xorg7_64 binaries).
Jonathan
by
jlisinski
-
AFNI Message Board
Hi - I used the AC-PC alignment (Define Markers), but I do not have the WARP_DATA entry in the header for some reason. I will try to figure out why. Thank you very much for your help with this!
by
jlisinski
-
AFNI Message Board
The AC-PC alignment was done manually in AFNI. I would not ask this question here otherwise. We do almost everything in AFNI if possible :).
by
jlisinski
-
AFNI Message Board
Hi -
Thanks for the quick reply! Unfortunately, none of these entries (WARP_DATA, ROTATE_MATVEC_000000) are in the .HEAD. I was also hoping to find the WARP_DATA entry. The only matrices that I found are: IJK_TO_DICOM and IJK_TO_DICOM_REAL. I copied one dataset here:
Thank you!
Jonathan
by
jlisinski
-
AFNI Message Board
Hi All,
I have an anatomical T1 dataset that has been manually AC-PC aligned and further processed (manually skull-stripped, etc). I would like to keep the processing, but undo the AC-PC alignment. I've looked at cat_matvec examples to see if I can extract the transformation matrix from the header and apply it/(the inverse) to undo the AC-PC alignment, but am not sure how to proceed. Any
by
jlisinski
-
AFNI Message Board
This variable was not in my .afnirc file, but I haven't checked README.environment. Thank you!
by
jlisinski
-
AFNI Message Board
This is great. There are still goodies in the controller that I haven't seen after many years :).
Is it possible to save 2. (Choose Zero Color) as a preference somewhere (.afnirc, layout, etc.) so it can be set in scripts?
by
jlisinski
-
AFNI Message Board
Thanks for your reply! My problem is finding x y z given the DICOM file. Any suggestions/ideas how you would do this? There is a link to a DICOM file in my original post.
Thanks again,
Jonathan
by
jlisinski
-
AFNI Message Board
Hi guys,
I have single-voxel MR spectroscopy data (DICOM, Siemens) and would like to know where the voxel (VOI) is located (create a mask for that single voxel and overlay that mask on a T1). Is there a to3d/Dimon hack to create a mask for the VOI given a DICOM file or write the FID into it?
Any help is appreciated!
Thanks!
Jonathan
by
jlisinski
-
AFNI Message Board
Hi,
1. I would use the same number of subjects/training samples for each group so that chance accuracy is the same across groups. However, you can also correct for bias if you are using an unequal number of training samples across groups.
2. The -predictions option will give you the distance to the hyperplane for each binary classifier <prefix>_<class label A>_<class label B
by
jlisinski
-
AFNI Message Board
Hi Nan,
Thanks for running your test again and I'm glad to hear that the classifier output does no longer differ when performing training and testing multiple times using the same command-line options. Most people probably would not have noticed, since these differences are smaller than single-precision. I will take a look at the multi-class predictions (DAG, vote) and check why the outp
by
jlisinski
-
AFNI Message Board
Hi Nan,
3dsvm was originally developed for classification of temporal and minimally pre-processed FMRI data. Thus, temporal de-trending of the classifier output is performed by default to correct for temporal drift that we observed. Please take a look at LaConte, 2007 (Fig. 4) for more details. However, performing de-trending of the classifier output is not the only and might not be the best w
by
jlisinski
-
AFNI Message Board
We were writing too many significant digits into prediction output files. I think this is the reason for the differences you saw. I updated 3dsvm and ran a similar test using motor data. All prediction files were identical regardless if executed sequentially or in parallel.
If you still see differences using the updated version of 3dsvm, please let me know.
Thanks for your help!
Jonath
by
jlisinski
-
AFNI Message Board
Hi Nan,
Thanks for running this and for letting us know! As long as model file and test options are the same, the distance to the hyper-plane for each observation in the predictions file(s) should be within single-float precision. However, I'm not sure why the number of different files is not roughly the same between sequential and parallel execution (chance ?!). I will run something si
by
jlisinski
-
AFNI Message Board
Hi Nan,
I added the option: -nomodelfile for this. This allows you to omit the -model option for training and no model file will be written to disc. I think this should be compiled overnight and will be available for download tomorrow ET.
Give it a try and let me know if you run into any problems.
Jonathan
by
jlisinski
-
AFNI Message Board
Hi Juan,
Yes, you can use this:
line 1: 9999
line 2-41: 1
line 42-80: 2
line 81-125: 9999
line 126-166: 9999
line 167: 9999
to train a single binary classifier (1-vs-2).
If you use this:
line 1: 9999
line 2-41: 1
line 42-80: 2
line 81-125: 0
line 126-166: 0
line 167: 9999
three binary classifiers will be trained and saved in the model (0-vs-1, 0-vs-2, 1-vs-
by
jlisinski
-
AFNI Message Board
Sure, no problem, I'm happy to help.
Yes, you can simply define the 1-vs-2 classifier using the -trainlables file and you don't need to label the sub-bricks belonging to the other classes with 9999s to exclude them from being used to compute the 1-vs-2 classifier. However, the computation time will be much longer if you include all classes, since 3dsvm will train classifiers for all
by
jlisinski
-
AFNI Message Board
Hi Juan,
I'm one of the developers and currently maintain 3dsvm in Stephen LaConte's lab.
3dsvm was originally designed for blocked designs, but it can be used for event-related designs with some extra prep work. We are planning to add options to better accommodate event-related designs in the future.
Questions, Run1_IM+tlrc & -trainlabels file:
3dDeconvolve -stim_time
by
jlisinski
-
AFNI Message Board
Hi Uday,
Your interpretation of the SVM weights is correct as well. If (label class A) > (label class B), weights contributing to class A will be positive and weights contributing to class B will be negative.
Jonathan
by
jlisinski
-
AFNI Message Board
Thanks for the reply Gang. I agree with you, i.e. "there is no way to say anything where you don't have any data". However, in this case, fdrval returns the t-value of the voxel with the smallest FDR q-value, which is somewhat misleading in my opinion. I think, it should rather return an empty string and/or an error statement.
Jonathan
by
jlisinski
-
AFNI Message Board
Hi guys,
I have a quick question about fdrval:
I generated a brik dataset using 3dANOVA2 and trying to compute a t-value given a q-value using fdrval:
fdrval -qinput results_anova+tlrc 1 0.01
The sub-brik for the t-stat is 1 and the input for the q-value is 0.01. fdrval returns 5.144 and this value does not change even if the q-value is set to 0.001.
Looking at the AFNI GUI, there
by
jlisinski
-
AFNI Message Board