Dear SPMers,

I have done two sample t-test, where in the first group there are (spect images of) controls and in the second group depression patients.

Using contrasts (-1 1) and (1 -1) I get clusters of hyperperfusion and hypoperfusion respectively. 

The problem appears when I want to check the cluster signal in individual images... In the SPMresults I mark a significant cluster from the list and then choose 'VOI -> cluster' and I get a graph of 1st eigenvariate against scan. As I understand in the scans on the axis are named 1-N in the order of entering them in design (first group, then the second group).

In the case of hypoperfusion the 1st eigenvariate is higher in controls (= left part of the graph is higher than the right). But in the case of hyperperfusion the 1st eigenvariate is again higher in controls, although the different contrast was used and the graph is drawn for a cluster that is significant for that contrast.

Where am I wrong?
I would appreciate any help, since I am confused...

Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen! 
Ideal für Modem und ISDN: