Skip to content

Latest commit

 

History

History
33 lines (24 loc) · 1.39 KB

preproc_plan.md

File metadata and controls

33 lines (24 loc) · 1.39 KB
Raw

Raw data location:

/data01/cas/EmotionInsula7T_project/Data_Collection

Anatomical

  • Rotate to std fslreorient2std
  • Inhomogeneity correction: using fast, but it's pretty bad inhomogeneities here - might need to use ANTs
  • Defacing apparently not needed
  • Skull stripping: again we need to see how fsl is doing here - might need ANTs

Functional

NB: it is important to separate the preproc from the stats stage in FEAT since it is likely that we will attempt many types of stats, while the preproc is always the same.

  • For each run
  • Reorient to std
  • topup with fmap
  • standard preprocessing w/out slice timing correction and spatial smoothing
  • registration nonlinear

Prepare predictors for stats

Details on how to build the EVs on the Feat guide

Note that since we have already carried out preproc, the input to the stats will be the .feat dir containing the preproc

  • triplets for each movie: onset (seconds), duration (seconds), intensity. Note that the first volume is at t=0

  • for a first-pass analysis, we will run four three-levels models (single run, second-level at the subject level, across subjects):

    • all movies
    • emotion rating (max across all emotions?)
    • arousal (emotion vs neutral?)
    • valence (negative vs positive?)

  • prepare EV.txt files in R. Hopefully also the contrast files can be prepared in R.