Dear SPM experts,   I am looking for an understandable approach by which one can check the realignment results.   my first question is if the starting point for the rigid registration will change by redoing the realignment on the same data? Lets say that we would like to register source image A on the reference image B; by the first try, the starting point would be x0. If we redo the same procedure after a day, will the starting point remain the same or will change?   The second question is  how can we appropriately check if realignment did a good job or not? It is extensively suggested that one of the best way is by visual inspection using check reg function in SPM, but can anybody let me know how we can do that(preferable an example which is provided somewhere)?    One of the crucial problems of minimizing the cost functions of mean square differences during optimization procedure might happen either when there is no global minimum or getting caught in local minimum rather than global minimum.Can any body let me now how we can check if this is the case or not? How can we change the starting point to get ride of this problems? How can we choose the starting point as close as possible to the global minimum?   It would be nice if one can give me one published examples of this issues as well.   Thanks in advance. /Kambiz rakhshan