Global-Task Correlations

calls sm_gltaskcorr
Computes the correlations between the different task 
conditions and the global brain mean intensity.

This function computes the correlation between the global variate (mean
activation in each scan) and the design conditions (columns of the design
matrix), with possibility to collapse over different sessions.

The user is prompted to specify the design matrix and images files for 
each session. The user should also specify whether columns corresponging to 
the same condition in different sessions should be summed over before
calculating the correlations. If the user chooses not to collapse, each 
condition column in the design matrix will be correlated with the global 
mean activation for only within the session which corresponds to that column.
If collapse is enabled, the accumulated column for each condition is correlated
with the entire global mean (i.e. the global mean for the session-concatenated 
timeseries).

The correlation coefficients are calculated and displayed for each task 
condition.

The utility assumes that the specified SPM design matrix contains the
following:
SPM.Sess(i).U(j).name - the name of condition j in session i (conditions
with identical names will be collapsed across sessions if collapse=1).
SPM.Sess(i).col - the columns in the design matrix corresponding to
session i.
SPM.Sess(i).row - the rows in the design matrix corresponding to session i.
SPM.xX.name - the name of each column in the design matrix
SPM.xX.X - the design matrix

This utility works with both SPM2 and SPM5.
Shay Mozes 04/2007
