Thanks Steve. I have a stage3 output map that shows statistical significance (p<0.05) and I'd like to use it to calculate the increase in co-activation for that component (the motor strip in my case). Subjects are scanned pre and post stimulus. I have the stage2 maps for the component in question, lets call it IC_1, for all 15 subjects, pre and post stimulus. They're in the file: dr_stage2_ic0001, a 4D image of length 30 (15 subjects * 2) I'd like to do the following: fslsplit this 4D image to get each subjects' map. divide each subjects' post map by their pre map to get a ratio map. average all 15 ratio maps -> mean_ratio_map use the stage3 map as a mask on the mean_ratio_map, and take the average over that masked region. This mean, I hope, represents the average increase in co-activation post stimulus for that component. Is this similar to what was done in the Filippini paper? Thanks.