I finished group ICA, and obtained group level components.
Now I want to reconstruct subject-specific (individual) spatial/temporal maps for each components, and test for correlation or group difference.
1. Can I extract one specific group-level IC (ex: DMN), and perform dual regression? or do I have to put all the components together?
2. I want to get individual map for each component using dual regression, and then perform statistical tests with the maps using matlab.
Is it okay to set design matrix as following? They are all nuisance covariates (age, IQ, gender), and I want to remove those effects.
group
average age IQ gender
1 35 109 1
1 25 112 0
1 22 102 0
1 29 110 1
1 26 121 0
average age IQ gender
C1 0 0 0 0
Any comment would be appreciated.
Thanks,
YM huh.
|