The Randomise program was run on the merged 4D image previously created, and the output prefix was set to TBI vs OI comparison.  The template brain mask was chosen to eliminate extraneous voxels from within the field of view, and the recently created design and contrast matrices were also chosen.  The program is set to run 5,000 random permutations for statistical significance, and the -T option is used to control for these many comparisons.