Hi Dmitry,
It seems very unlikely that this has anything to do with
the re-ordering of your datafiles. In fact, you might
just look at them without renaming the files. I suspect
you will still see the problem.
This looks like it could be an aliasing artifact from
the scanner, though the garbage at the top of the image
is suspicious. You should make sure that you are really
getting complete head coverage when scanning someone
(and that the middle of their head is near the 0,0,0
coordinate).
If you view this data in afni, and put the crosshairs
at the 0,0,0 coordinate, are you in the anterior part
of the subjects brain?
- rick