The physiology studies are loaded separately, using their original sampling rates.
The fMRI-specific physiology is linked to fMRI studies in the Project configuration.
The question (2022-08-08) is happens next. Let’s consider different scenarios:
Scenario | Output | Preprocessing and parameters |
|---|---|---|
View concurrent physiology | Time-trends with RMANOVA | Same as physiology Time-points for bins: TR (same as fMRI) |
Corelate whole-brain fMRI | Correlation maps of BOLD*phys This should be average of phys channel over TR collection period of fMRI scan | Time-points for bins: TR (same as fMRI) Get one value/TR time-point Correlation that series with each voxel time-course ?Convolve with hrf? What preprocessing? Motion correction? Do normalization steps on correlation maps Provide correlation maps for 2nd level analysis |
Slice-by-slice correlation | Correlations maps Categorize location in breath: inspiration, expiration, end-insp, end-exp (and so on) | Get time-bins for slices, get phys for those time-bins Create correlation maps on a slice-by-slice basis ?How to deal with motion correction? Create phys time-trend for each voxel (slice) |