Hi, In my experiment I have two runs and I noticed that when I display the original functional img images the origin is very different (in some subjects) for run1 and run2. What is the correct way to reorient images in order to set origin as correct as possible to the AC: A- to determine the position of AC in the first functional image of run 1 and apply reorient images to all images (run1 + run2) B- to determine the position of AC in the first functional image of run 1 and apply only to run 1 and then determine the correction from the first image of run 2 and apply to run2. I am asking this because initially I was doing the option A and I found that in some cases the correction for run 1 is not good enough for run 2, and the origin continues to look distant from AC. Thanks, Mirtes