[B5I6] Pipeline reproduction (FSL, deriv)

[mirhnius]

Softwares

FSL 5.0.9

Input data

derivatives (fMRIprep)

Additional context

(copied from spreadsheet)

1. Data received in fmriprep pre-processed form; pre-processing steps performed in order specified in fmriprep v1.1.4
2. "FramewiseDisplacement" column extracted from the fmriprep confounds file for each run, plotted and used to identify volume-to-volume displacement
3. Extracted 6 columns from fmriprep confounds file for each run (X, Y, Z, RotX, RotY, RotZ) and computed square, temporal derivative, and temporal derivative squared to create 24-column motion confound file (in Python), which were used to plot volume-to-volume displacement for each of a subject's 4 runs

List of tasks

Please tick the boxes below once the corresponding task is finished. 👍

• 👌 A maintainer of the project approved the issue, by assigning a 🏁status: ready for dev label to it.
• 🌳 Create a branch on your fork to start the reproduction.
• 🌅 Create a file team_{team_id}.py inside the narps_open/pipelines/ directory. You can use a file inside narps_open/pipelines/templates as a template if needed.
• 📥 Create a pull request as soon as you completed the previous task.
• 🧠 Write the code for the pipeline, using Nipype and the file architecture described in docs/pipelines.md.
• 📘 Make sure your code is documented enough.
• 🐍 Make sure your code is explicit and conforms with PEP8.
• 🔬 Create tests for your pipeline. You can use files in tests/pipelines/test_team_* as examples.
• 🔬 Make sure your code passes all the tests you created (see docs/testing.md).

[bclenet]

Hi @cmaumet,

In the description, under independent_vars_first_level :

Other nuisance regressors: 1 column for each volume identified as an outlier (outliers identified based on “non-stdDVARS” and “FramewiseDisplacement” columns from fmriprep confounds files (threshold at 75th percentile + 1.5 times interquartile range)). Each column had value 1 at the timepoint identified as an outlier, and zero at all other timepoints.

What I understand from that is :

1. compute outliers of the non-stdDVARS column according to the formula
2. compute outliers of the FramewiseDisplacement column according to the formula
3. create a regressor from step 1 and 2 (OR operation between the two)
4. add the regressor in the model, only if there is at least one 1 in the column

I just wanted to be sure it's the right thing to do.
Thanks !

[cmaumet]

@bclenet that's perfectly how I understand it as well!