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 2251 - 2280 of 2305
Hi,
I am interested in regressing out local WM from some resting data, which I see is done with `-regress_anaticor' in afni_proc.py. In the helpfile example 9b, CSF is not additionally regressed, I notice, even though WM+CSF regression often go together-- is this a subtle hint that such would be a bad idea, for example overkill of regression? Or is that something which is still not blat
by
ptaylor
-
AFNI Message Board
Hello,
A pre-print of the paper describing the AFNI tracking+connectivity tools, "FATCAT: (an efficient) Functional And Tractographic Connectivity Analysis Toolbox" by Taylor & Saad (2013, to appear in Brain Connectivity), is now available, located in the FATCAT section here:
along with the OHBM poster and the paper (Taylor et al. 2012) describing the underlying tracking proce
by
ptaylor
-
AFNI Message Board
Hello,
There is a newly available demo of the FATCAT (Functional And Tractographic Connectivity Analysis Toolbox) v1.1 subset of programs available for downloading/running; it can be got in the same way as the old toolbox, by typing the following in the commandline:
$ @Install_FATCAT_demo
which will download and unzip the FATCAT_DEMO/ directory with all necessary files and scripts (NB: it
by
ptaylor
-
AFNI Message Board
Hi,
The bug-fixed version is now available in the latest AFNI download (the hot-off-the-press one today). Please let me know if errors persist.
Also, Laura, I notice that the number of ROIs in each of your networks appears to be 1. I guess that means that the largest integer in your targer masks in each case is unity. If you are using probabilistic tracking and wanting to look for AND log
by
ptaylor
-
AFNI Message Board
Hi-
The reason for both of these `error' messages is that there is a small bug in the program about dealing with RAI files. I have sent a fixed up version to be put into the newest source code, which should be in quite soon.
Thanks,
pt
by
ptaylor
-
AFNI Message Board
The FATCAT (Functional And Tractographic Connectivity Analysis Toolbox) demo set is now uploaded online and available. In the most recent AFNI source code there is an @Install_FATCAT_DEMO script which will download and unzip the directory. One can execute it by typing the following in a command line:
$ @Install_FATCAT_DEMO
and a directory FATCAT_DEMO will be made and filled, there and then.
by
ptaylor
-
AFNI Message Board
Hi, Piero-
The new demo actually hasn't been put in place of the old one yet, but it should be soon...
If you'd like one sooner than that, I can share a copy of it if you have a gmail account (it's about a 200MB folder). You can email me (neon.taylor@gmail.com) if so.
--pt
by
ptaylor
-
AFNI Message Board
Ah, you're right, that's the older demo files... I'll have to see about the newer ones.
--pt
by
ptaylor
-
AFNI Message Board
The demo data and scripts can be obtained by typing the following into a terminal:
$ @Install_PTaylor_TractDemo
--pt
by
ptaylor
-
AFNI Message Board
Well, some assumptions go into making and using the tensor model of diffusion, such as symmetry of measures about the origin and a Fick's Law-based or probabilistic interpretation of diffusing particles (particularly about deriving the latter, see, e.g., Minati & Weglarz 2007). I think that for the tensor fit to be interpreted as the surface of an ellipsoid, mathematically the tensor/ma
by
ptaylor
-
AFNI Message Board
Hmm, ok, I understand better now about the precision required. I'm curious what differences are found between dtifit and 3dDWItoDT? Whenever I've looked at both/either, I've seen them give quite similar results, except for the `bad' voxels which might suffer a lot from partial voluming at ventricles, etc. (In those cases, I think 3dDWItoDT and dtifit differ in how they deal
by
ptaylor
-
AFNI Message Board
Hola-
I am a bit curious about why eight sig-figs would be needed for these numbers? I can't imagine that the DTI estimations (due to scanner noise, eddy, motion, eeetc.) could allow those values to actually *be* significant to that accuracy-- one has to remember that there should be something like +/-standard deviations or confidence intervals on estimates (even though those aren't
by
ptaylor
-
AFNI Message Board
It's in units of `indices' (so you have to use the edge length of voxels to change to think about mm).
--pt
by
ptaylor
-
AFNI Message Board
The new version of 3dReHo is in the current AFNI build, with flexible neighborhood sizes and shapes.
Please let me know of any further suggestions to improve this and the RSFC programs, but I am sure that Ziad will be happier if any new questions about Reho/RSFC calculations start on a new thread....
Thanks,
pt
by
ptaylor
-
AFNI Message Board
Hi,
Ok, I updated 3dReHo, and I hear from the powers that be that it should be in the next AFNI build (~next couple days).
In it, you can now choose a radius R of any size in terms of numbers of voxels in order to make a `discretized sphere' neighborhood of voxels, each of whose index is <=R from the central one. Some examples of radii, physical volume (V=[4*pi*R^3]/3 in arbitrar
by
ptaylor
-
AFNI Message Board
Hi,
That seems to make sense to me; I believe that the errts files would be the right ones-- perhaps Rick could say definitively about that.
--pt
by
ptaylor
-
AFNI Message Board
Hi,
Well, there is an option there currently for calculating ReHo/KCC for ROIs. Would you want every voxel though to have a larger neighborhood? For example, right now, the voxelwise neighborhood is a 3x3x3 centred cube (from which you can select symmetric subsets). Would you like to be able to calculate over a 5x5x5 or 7x7x7 or other such cube volume per voxel?
--pt
by
ptaylor
-
AFNI Message Board
At the moment, my thought would be that processing with afni_proc.py is a good thing in terms of the overall data; the ability to include the RSFC parameter estimations directly in this would be useful, yes.
I don't think one can use 3dRSFC on data which has already been LFF-ified, since parameters like fALFF will need un-bandpassed spectra for comparison.
In afni_proc, I believe that
by
ptaylor
-
AFNI Message Board
As Ziad correctly pointed out, 3dMatch can do the job of calculating both (spatial) correlation or the Dice coefficient-- and I don't think it will even be thaaat tortuous to use. If you're getting zeros from subtraction, I'll assume that you are wanting to compare to binarized/integerized maps? That can be done with 3dMatch as well; you can even specific a min/max range of valu
by
ptaylor
-
AFNI Message Board
Hi-
I'm wondering what is meant by 'creates a large amount of blur'? ReHo is MRI-lingo for Kendall's coefficient of concordance, and it estimates a measure of similarity among a group of time series (based on rank values within each time series). In the most common case, ReHo at a voxel with index (i,j,k) involves a calculation with voxels at: (i-1,j-1,k-1), (i-1,j-1,k),
by
ptaylor
-
AFNI Message Board
Ok, switching off the version check seemed to work. Hopefully, that means I haven't been blacklisted...
--pt
by
ptaylor
-
AFNI Message Board
Hi,
Using the most recent AFNI update on my Mac OS 10.6, I get the following message every time I run 3dcalc:
Can't connect? tcp_connect: Connection refused
(and actually, it appears twice). This has not happened previously. Also, the calculation appears to be done fine. It's more of just an annoyance thing to see this-- it slows things down a bit waiting for the second coming
by
ptaylor
-
AFNI Message Board
There are some updates and new inclusions to the set of programs in
AFNI that have been written for assisting in combined studies of
tractography and FMRI. Ziad and I are working on an expanded set of
demo files and a more formal presentation of their applications, but
in the meantime here are brief descriptions of newest changes and
additions:
3dMatch: (new)
Input two files (probably
by
ptaylor
-
AFNI Message Board
Hi-
There are voxel values of ALFF, fALFF and mALFF output by the code. ALFF is just the sum of spectral amplitudes of LFFs at a voxel; mALFF is a voxel's ALFF divided by the average ALFF in the brain (I think this is the one you are asking about); fALFF is the voxel's ALFF divided by the same voxel's total (non-LFF filtered) spectral amplitude.
--pt
by
ptaylor
-
AFNI Message Board
Well, by looking at it, I would venture a guess that it is an 'ideal filter', meaning that it has a binary modulation transfer function (equal to 1 in the bandpass region and 0 outside of it). That is, frequencies in the bandpass range go through the filter with full amplitude, and any unfortunate souls not in that interval get squashed to zero. (Though, I think at a minimum, the zer
by
ptaylor
-
AFNI Message Board
Hi, JoJo-
That certainly does not sound impossible. You happy with the 3dROIstats output format? Or would you want just a list just the numbers: a column of numbers, length determined by number of ROIs?
Cheers,
pt
by
ptaylor
-
AFNI Message Board
Hi, Rick--
Yep, you beat me to the reply, but yes, instead of leaving in a 'non-BP' bandpass, I just deleted that line, and it worked fine through to the end.
--pt
by
ptaylor
-
AFNI Message Board
.... that being said, I get a failure message in the list of results at the end:
...
++ 3dOverlap: AFNI version=AFNI_2011_12_21_1014 (Oct 26 2012) [64-bit]
#A=./full_mask.Y09+orig.BRIK B=./mask_anat.Y09+orig.BRIK
#A #B #(A uni B) #(A int B) #(A \ B) #(B \ A) %(A \ B) %(B \ A) Rx(B/A) Ry(B/A) Rz(B/A)
61255 51971 64105 49121
by
ptaylor
-
AFNI Message Board
Ok, thanks. I had deleted the line manually in the actual file, saved, and then tried to run, but that hadn't worked. Making the changes in the pop-up window, saving, and then running did work fine.
Thanks much,
pt
by
ptaylor
-
AFNI Message Board
Ok, I'm doing something wrong (I admit, this is my first time uber_subject'ing).
"show proc script" yields the following:
** warning: removing first 5 TRs from beginning of each run
--> the stimulus timing files must reflect the removal of these TRs
** cannot warp to tlrc space without +tlrc anat via
either -copy_anat or the 'tlrc' processing block
by
ptaylor
-
AFNI Message Board