I need a LOT more information about what you are doing.

What "parameter" are you talking about? Are you working
on anatomical/structural images or a functional series, or
have you tried it on both? What makes you think there is
an error?

to3d should work, although it will not preserve the obliqueness
information properly. It should at least generate viewable
images, and if all your images are aligned the same, off the
top of my head I'd say most statistical analyses shouldn't
be affected.

There are some specific problems with Siemens and Phillips
scanners we've identified that cause problems, so there
might be something with GE. But I need more information.

Perhaps you could begin by answering the questions I
asked above, and also sending me the command line
you are using for to3d, and the output of that command.