I have carefully gone through the tutorial data "good_output", and find that wm.mgz lies inside the boundary of the white matter almost all the time, with fewer than 10 slices of exception. Whenever there is an exception, the affected area is relatively small. Therefore, I am wondering

1) Should wm.mgz always lie inside the boundary of the white matter?

2) If I have subjects whose wm.mgz lies outside the boundary of the white matter (which I do), should I add control points until the two nearly match?

Thank you very much!

Sincerely,

Ye