I am new to resting state analyses, and I would like to complete the analysis using afni_proc.py and the recommended FreeSurfer segmentation methods from the Jo et al., 2010, NeuroImage article. I am confused, however, about an aspect of the implementation.
Based on afni_proc.py -help’s example 11 – it looks like anaticor processes occur within the afni_proc script (after the user runs FreeSurfer segmentation on the subject’s data). However based on Hang Joon Jo’s technical notes page (https://afni.nimh.nih.gov/sscc/hjj/anaticor) and the @ANATICOR –help page, it outlines in preparation for anaticor to first complete EPI preprocessing, then use the output to compute FreeSurfer segmentation, then to run @ANATICOR. I am confused about where afni_proc.py example 11 comes in…
Do the steps go something like this:
Run afni_proc.py without the anaticor options to complete initial EPI preprocessing
Complete FS segmentation
Rerun Afni_proc.py with the anaticor options e.g., -regress_ anaticor_fast and the FS segmentation files
Where does running @ANATICOR come in? - or is that taken care of in afni_proc.py with the anaticor options (e.g., as outlined in example 11)?
Please review “FREESURFER NOTE:” in the afni_proc.py -help
output. It shows exactly what commands were run (just a
few) to generate the FreeSurfer inputs needed for example 11.
FreeSurfer is run first.
The main point of @ANATICOR is the local white matter
regression, applied in afni_proc.py via -regress_anaticor
or -regress_anaticor_fast.
Two followup questions in this vein - in the afni_proc.py resting state example 11 with anaticor and FreeSurfer based tissue regressors, the tlrc_base is specified to MNI_caez_N27+tlrc - is this just for illustrative purposes on how to use this option or is this a recommended base?
Also, the align_e2a option is specified in this example (and many of the previous resting state examples) - again, is it recommended to align the EPI to the anat for resting state with anaticor and FS based tissue regressors or was this for illustrative purposes? If it was illustrative, are there recommendations on going one direction vs. the other?
There is no recommended template, as there is not even a
recommended space. First choose an atlas that suits
whatever atlas needs you may have, if there is one. Then
choose a good template that is aligned with the atlas.
What makes a template good? Optimally it is made from
many similar subjects aligned via non-linear registration.
Note that MNI_caez_N27+tlrc is just one subject. It is
not optimal, but it is still a good template (IF you want
to be in MNI space).
Really, I would expect e2a to make almost no difference
at all, since there is a subsequent registration with a
template. It is much more important when staying in
orig space. Okay, I will run a tiny test, just to see…
By the way, thank you for understanding and pointing out
that some options in examples are actually just for
illustrative purposes! I may have to take most such
options out (I have started), as they can be confusing
(since they may seem recommended).
I’ve finished the resting state preprocessing steps and I’m now getting ready to create my correlation maps with ROI masks, 3dmaskave and 3dfim+. I wanted to verify that I am extracting information from the correct input file though. I see that the output of afni_proc.py with ANATICOR has two errts files:
I’m having trouble locating documentation on the difference between these two files however - could you point me in the right direction? Is one recommended over the other for resting state analyses?
Also is there a central location in which the various output files from afni_proc.py are explained?
To some degree the documentation for the output is
the proc script itself…
In your case, errts.tproject was presumably made
before errts.fanaticor. They are basically the
same, except that the fanaticor version includes
application of the ANATICOR regressor.
rick
The
National Institute of Mental Health (NIMH) is part of the National Institutes of
Health (NIH), a component of the U.S. Department of Health and Human
Services.
I may have to take most such