functional connectivity differences using SwE

[Alon Itzkovitch]

I have a resting state data of two groups with five timepoints. I tried to measure the differences between the functional connectivity of the groups using the SwE toolbox.

First, I computed a functional connectivity matrix for each scan (subject X in session Y). Next, I convert the connectivity matrices into one matrix (meaning each row is a specific scan, and the columns are the edges).

For example, it looks like the extended version of this:

egde3	egde2	edge1
0.5	0.6	0.7	sub-1_ses-1
0.4	0.4	0.5	sub-1_ses-2

For the design matrix, I used different columns for the combination of group and time:

group1_session1  group1_session2 .... group2_session5

Finally, I used the wanted contrasts:

1. main group effect: 1 1 1 1 1 -1 -1 -1 -1 -1

2. main time effect: -2 -1 0 1 2 -2 -1 0 1 2

3. interaction between group and time: -2 -1 0 1 2 2 1 0 -1 -2

My question is:

I want to make some statistical interpretation of the results I got, but I'm not sure how to do so. I have the z-score matrix but am I need to threshold it? and how?

Maybe I need to use another file (like the p-value results) for this kind of analysis?

Thanks a lot!

Alon