Dear all,
I have some questions concerning ROI (FFA, FBA and EBA) analysis and preprocessing. I have separate localiser data that I use for defining the ROI's.
a. Localizer= Partial Brain
b. Normal Experiment= Whole Brain
My plan so far:
1a. Display: Reorient Whole Brain to AC-PC line
b. Open 2nd display window and get the same orientation for partial brain as whole brain (just by comparing visually)
->This step takes a lot of time because I have many subjects. I tried in one participant to skip this step and looked at the images after normalisation. I don't see a difference between the reoriented and the ones where I started slice timing without reorienting. But maybe I should compare after first level analysis. What do you think? Is it safe to skip reorientation in this case? And how can I best check if it is?
2. Slice Timing:
a. Whole Brain
b. Partial Brain
3. Realignment:
a. Whole Brain
b. Partial Brain
4. Coregister:
b. Partial Brain with Whole Brain (Target: Mean of Whole Brain, Source: Mean of Partial Brain, Other Images: Realigned Scans of Partial Brain)
5. Spatial Normalization:
a. Determine Normalization Parameters (Target:EPI.mnc, Source: Mean of Whole Brain)
b. Write Normalized (Realigned Whole Brain images, Coregistered Partial Brain images)
-> Do I actually need normalisation when I want to define ROIs? Or is coregistration enough?
6. Smoothing
a. Normalized Whole Brain images and Normalized Partial Brain images
->Is smoothing recommended with ROI analysis or better skip this step?
->I want to pick in each subject the most significant voxel, put a sphere around it and only include the significant voxels that are connected with each other.
I have no idea how to do that in SPM5 and whether this is the best method.
Any suggestions are welcome!
Thanks a lot,
Mariska