Are they CT images? These often have a shear in them.
MR images should not be sheared. Usually, the voxel to world mapping in the
headers can be represented by nine parameters (translations, rotations and
orthogonal zooms - in a particular order). For some reason, the mapping in
your header can not be exactly represented by these nine parameters. Maybe
there is some small rounding error when SPM checks if the header can be
represented this way. If the precision with which the image orientations are
stored in DICOM is not high enough, then SPM may detect a very slight
deviation - which is probably nothing to worry about.
Best regards,
-John
On Tuesday 19 August 2008 21:19, John Gelburg wrote:
> But I didn't make any zoom explicitly; I just displayed the image.
> Might it be that SPM5 zoomed the image automatically for me? Are
> there any consequences for my analysis or this is the matter of
> display only?
> Thanks.
>
> On 8/18/08, John Ashburner <[log in to unmask]> wrote:
> > This is a warning that occurs when you zoom etc an image using the
> > Display button.
> >
> > Best regards,
> > -John
> >
> > On Friday 15 August 2008 21:37, John> <Gelburg wrote:
> >> Hi,
> >>
> >> I got this warning just at start, before any preprocessing (I just
> >> converted to NIFTI from DICOM).
> >> What are my steps to fix it?
> >> How harmful is it?
> >> Do I make something wrong during scanning?
> >>
> >> Thanks!
|