Hi John,

Getting a map of the average correlation with all
gray matter voxels is something an afni_proc.py
script almost does already. Instead, it computes
the average correlation of all pairs of "brain"
voxels (gray, white, CSF), calling that GCOR.

Looking at such a script, GCOR is computed (from
the errts dataset) via:

3dTnorm -norm2 -prefix rm.errts.unit errts.${subj}+tlrc
3dmaskave -quiet -mask full_mask.$subj+tlrc rm.errts.unit+tlrc > gmean.errts.unit.1D
3dTstat -sos -prefix - gmean.errts.unit.1D\' > out.gcor.1D

To alter this:
1. Change the input to 3dTnorm to whatever is
appropriate, or maybe the same thing.

2. Input a gray matter mask to 3dmaskave.

3. Instead of the 3dTstat command, correlate
each voxel with the result from 3dmaskave.

3dTcorr1D -prefix gray.corr rm.errts.unit+tlrc gmean.errts.unit.1D

This will produce a whole volume map, not just
restricted to the gray matter. You can further
restrict the result if desired.

- rick