a basic question about alff

Dear Dr. Yan,

            Sorry to bother you some basic questions, I want to know what I will get after two-sample t-test by using ALFF, the data I uesd mALFF. Second, if I used VOI in spm to extract significant value from two-sample t-test by using ALFF, what's the value represent? Third, if I have used covariates such as gender in the two-sample t-test, after that I extracted value to do correlation, should I use partial correlation, I mean regress out gender covariates again? or only correlation?

Any helpful advice will be highly appreciated!



Hi Fish,

1. This is the difference of normalized ALFF (m*) between the two groups.

2. If you extracted from the m* files, that means the relative strength of ALFF of that voxel (relative to the global mean).

3. I suggest you to use a GLM model to do that. Partial correlation is equivalent to regressing out all the other covariates and then do correlation.