I'm running a functional group analysis using some preexisting data. After running through all of the data, I noticed some scans look a bit tilted from the first level post-stats results. (Think a 30 to 40 degree axial tilt from midline). Registration looks perfect, but I was wondering if the tilt factors into the group analysis?

If everyone's registered to standard MNI space, does it matter if the scan FOV was tilted such that the volumes acquired are slightly tilted? Or should I automatically remove these subject brains from my analysis?

Cheers,

Kevin Japardi