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 - 23 of 23
Hi Rick,
yes, absolutely - would be great if you could have a look!
I actually still had the problem when replacing all NAN by 0 before writing the file...
Best,
Janina
by
janina_aletheia
-
AFNI Message Board
Hi Rick!
No, I havn't used AFNIio.R as far as I am aware, just oro.nifti - is that better?
3dcalc seems to sort of fix it: now the OLay is -6 to 5 which seems the rounded actual min and max values (-6.270428 and 5.36703). Do you think it is correct like this now?
Many thanks!
Janina
by
janina_aletheia
-
AFNI Message Board
Hi,
I run a voxelwise regression analysis in R and saved the Z-values in an 61*73*61 array and exported it as nifti using oro.nifti. This numeric array contains also NANs but the numbers range from -5.65 to 6.44382. However, in order to get a file that could be read, I used the header of a single subject seed correlation file (imported it as nifti object to R and replaced the data array while
by
janina_aletheia
-
AFNI Message Board
Hi Gang,
thanks for the answers!
>So you want to use SRS to differentiate the twins within each pair? You would have to make that judgment call since that seems to be > a science (instead of statistics) question.
Yes, that is the plan. My sample consists of a mix of different types of pairs: concordant or discordant for ASD, other NDD or TD. Therefore comparing groups based on diag
by
janina_aletheia
-
AFNI Message Board
Hi Gang,
thank you so much for the fast reply! I do have some follow-up questions. Maybe I am still lacking some basic understanding of what the model does - tried my best to understand but hope you can help!
>In that case it's probably better to not consider a fixed effect for twin. So maybe a simpler model like this? -model "SRS+age" \
I guess you are right, the main
by
janina_aletheia
-
AFNI Message Board
Dear Gang and other 3dLME experts,
I have a related question regarding using 3dLME for twin data as suggested in Chen et al. 2013 in example 6. The images are seed correlation Z-maps and the quantitative predictors are SRS-score as measure of autistic traits (per twin) and age (per pair). So I actually have only 1 measure per twin, but 2 measures per pair. I am mostly interested in the relatio
by
janina_aletheia
-
AFNI Message Board
Hi Rick,
good point! That's right, the ones exceeding the threshold are indeed on average younger. But I would lose most of the very young ones also with a more liberal threshold. Unless I would go even above .5 threshold - but then what is the point in censoring if only picking out the really big movements? Hard to find a good compromise.
Thank you for your input!
Janina
by
janina_aletheia
-
AFNI Message Board
Hi Rick,
I do understand that the loss of DoF is a problem, I was just wondering why it was used in example 9 but not 11 and whether the more elaborate nuisance regression would make it either problematic or not needed. After reading all those bp discussions now I don't feel too keen on using it at all any more, but I have been advised to use it and I also read here somewhere that it is s
by
janina_aletheia
-
AFNI Message Board
Sorry: and an additional question: why should one regress both, local and average WM? I thought that the local WM is a more accurate way of WM regression, making the global WM regression unneccessary.
In the technical note for ANATICOR:
It says:
So, the latest model we use is:
Y = [ MO + RI + WMeLOCAL + LVe ] X + Residual
I do not have physiology measures, so the RI regression is not pos
by
janina_aletheia
-
AFNI Message Board
Thank you very much, Rick! Seems much clearer now.
Best,
Janina
by
janina_aletheia
-
AFNI Message Board
Hi Rick,
thanks for the clear answer - I will have a closer look at example 11 then. Is saving degrees of freedom the only reason for not using bandpassing here or are there other problems when adding it to example 11? Does it cost many DoF to regress the first 3 principal components as compared to just regressing mean CSF signal?
When you say one would rather censor more: do you mean that
by
janina_aletheia
-
AFNI Message Board
Hi experts,
I would like to make use of the local WM regression from ANATICOR like in example 9b from anfi_proc.py in combination with regression the average eroded CSF like in example 10 for resting state preprocessing - please see below.
1) Does anything speak against this combination?
2) In example 11, it is not local but average WM, right?
3) How much better is the "first 3 prin
by
janina_aletheia
-
AFNI Message Board
Dear Rick,
Sorry for my persistant lack of understanding, but I have still difficulties with this topic. I read all I could find about it on the message board but there is still some confusion:
From:
You wrote:
“Censoring before bandpassing breaks the time axis in a
model that is all about signal frequencies. Censoring
after bandpassing can send echoes of would-be-censored
spikes ri
by
janina_aletheia
-
AFNI Message Board
Hi Rick,
oh, NOW I get it. I got confused because the errts.tproject file has empty volumes where the censor file is 0 while the errts.anaticor file does not and I didn't see that the amount of volumes was reduced also in the errts.anaticor file, just without leaving "gabs". I misunderstood the keep_TRs variable and assumed it means all TRs are kept (so ignoring the censoring).
by
janina_aletheia
-
AFNI Message Board
Thanks very much, Rick! Great to get all this support!
Janina
by
janina_aletheia
-
AFNI Message Board
Hi Rick,
Yes, I use afni_proc.py. I just checked: I initially did not bandpass and then added the pb just now: indeed without pb I use 18 DoF while with bp it is 107 (leaving 72 in a person where 21 volumes were censored). Wow, thanks for pointing this out, I had no idea bandpassing used so many DoF! So maybe I should skip that step?
Best,
Janina
by
janina_aletheia
-
AFNI Message Board
Hi,
I installed afni via the Ubuntu software center/NeuroDebian. It is running fine, but using afni_proc.py with Example 9b (anaticor and censoring) let to 2 datasets, one with local WM regression and one censored but I want both methods combined. I saw in another thread that anaticor now should work with 3dTproject but looking at the script created with afni_proc.py I see that in my case it i
by
janina_aletheia
-
AFNI Message Board
Hi Rick,
thanks for clarifying this, makes sense.
Is there actually a rule of thumb how many volumes need to be left for a reliable seed correlation? What other factors does it depend on? I read often that researchers exclude participants with less that 90 or 80% of data left but for me that would mean that I lose more than half. I have a 10 minutes RS scan leading to 200 volumes. Looking at
by
janina_aletheia
-
AFNI Message Board
Hi Rick,
thank you very much for your reply.
I thought that one concern might be that if there are too few volumes in a row, there might not be enough left of the temporal structure of the data. However I don't understand how that affects e.g. a seed correlation I perform on the censored data. And I do not know whether it is not actually worse for seed correlation to lose data points (
by
janina_aletheia
-
AFNI Message Board
Dear experts,
I read in recent publications on resting state data that the authors excluded also time points if they were not in a row of at least 10 consecutive usable time points. My questions:
1) Do you recommand this for resting state analysis? (and could you please explain why or why not)
2) If so, is 10 volumes a reasonable cutoff in your opinion?
3) Is there an option to specify th
by
janina_aletheia
-
AFNI Message Board
Dear Rick,
thank you very much for this fast and clear response, now I understand.
One related question:
It seems that when using the default otions, uber_subject.py creates a script that censors the TR exceeding the threshold plus the previous TR. I don't understand why I should censor the previous TR - I thought there are rather effects of motion on the following TRs from what I rea
by
janina_aletheia
-
AFNI Message Board
Hi,
I am just exploring AFNI as an alternative to SPM for resting state preprocessing, using the uber_subject.py and noticed that the estimates of head motion (saven in the motion_[...]_enorm file) is quite different from the framewise displacement calculated from spm motion parameters in matlab. Could someone clarify:
1) what exactly the difference is between FD and the motion estimate from
by
janina_aletheia
-
AFNI Message Board