Dear FSLers
I am wanting to run a PPI analysis using an ROI in standard MNI space on some task fMRI data which was performed in 2 runs and then combined in a fixed effects analysis. The data has already been run in an initial 1st level Feat without PPI including FNIRT registration.
As I understand it there would be 2 ways of doing this depending on whether one works in subject or standard space.
Which would be the more appropriate method?
1. Run PPI on fMRI data already transformed into standard space and use the ROI for time series extraction in standard space
a) Convert the individual filtered_func images from the initial (non-PPI) Feat into MNI space using applywarp
b) Extract time series from the filtered_func_mni images using the ROI in standard space
c) Run Feat on the filtered_func_mni images including the PSYC, PHYS and interaction term as EVs separately for Run 1 and Run 2
d) Combine Run 1 and Run 2 using fixed effects higher analysis
e) Run group level higher analyses
If this is the correct method what do I do for registration at stage (c) to enable there to be something to take up into step (d) given that the inputs are already in standard space?
Can I just include a single FLIRT registration step into standard space with just 3DOF registering the average_filtered_func_mni image to itself as image already in MNI space?
2. Run PPI on fMRI data in subject space using the ROI for time series extraction transformed into individual subject space.
a) Convert the ROI in MNI space into subject space using applywarp with parameters from an individual's initial (non-PPI) Feat.
b) Extract time series from the filtered_func images using the ROI in subject space
c) Run Feat on the filtered_func data from initial non-PPI Feat (to avoid having to redo motion correction) including the PSYC, PHYS and interaction terms as EVs separately for Run 1 and Run 2
d) Combine Run 1 and Run 2 using fixed effects higher analysis
e) Run group level higher analyses
If this is the better method how do I avoid having to completely redo the FNIRT registration step at stage (c)?
Can I just copy the reg folder from the initial (non-PPI) Feat into the new PPI Feat folder?
If so, do I then need to run Featregapply on the new PPI Feat folder before running the second level Fixed effects analysis to combine runs?
Many thanks for your assistance.
Regards
Tony
Dr Tony Goldstone
MRC Clinical Sciences Centre
Imperial College London
Hammersmith Hospital
|