Update QC recommendations after running fMRIPrep + partial afni_proc.py (mask/scale/regress only)

Hi AFNI Gurus,

I’m hoping to get some guidance on best practices for QC at a slightly nonstandard point in my pipeline.

Pipeline overview

I ran fMRIPrep for initial preprocessing and QC, using it for what I understand to be the functional equivalents of the following afni_proc.py blocks:

  • tcat
  • tshift
  • volreg
  • blur
  • align
  • despike
  • tlrc

After that, I ran afni_proc.py only for the final stages, specifically using just the following blocks:

  • mask
  • scale
  • regress

Because I only ran these later blocks in afni_proc.py, the usual QC HTML and driver scripts were not generated automatically.

I’d now like to do appropriate QC on the outputs from afni_proc.py at this stage and wanted to check whether the steps I’m planning are sufficient, or whether there are additional QC checks you’d recommend.


My current QC plan

Since I don’t have the auto-generated QC HTML, I’m planning to manually review the following:

1. General warnings / failures

  • output.proc.s##
    Read through the main script output log and search for any WARNING or FAILURE flags.

2. Alignment

  • out.mask_ae_overlap.txt
    Review EPI–anat alignment via overlap between EPI and anat brain masks.
    Check the %(A \ B) value and confirm it is not excessively large.

Hello,

I would not have expected fmriprep to do any blurring, and having the align and despike blocks where they are seems a little odd.

Assuming your AFNI version is very current, consider examples "AP demo 3a" and 3b and 3c. As seen with afni_proc.py -show_example_names, those are examples for use after fmriprep.

Importantly, you would want to pass the motion file to afni_proc.py, possibly as well as some confounds. Consider:

afni_proc.py -show_example 'demo 3b' -verb 2

Let me know how reasonable that looks to you.

-rick