[MINC-users] Computing a "orthogonalizing" rotation transform?

[Andrew Wood]

Hi Gabriel,

You might already be aware of this, but Nifti files can simultaneously
encode two different world spaces (see qform and sform).  It's my
understanding that the Nifti standard stately clearly how they're to be
used, but many toolsets choose to not conform; one must be very careful
when using tools coming from different nifti toolkits.

Here's an example where the qform/sform issue could lead to confusion: a
viewer could use the sform to position the image in the viewer, but then a
nifti-minc converter could pick up the qform to generate the world axes;
the sform orientation information would then be lost.

Not sure if that helps, but it could be related.

Cheers,
Andrew

On Wed, May 4, 2016 at 2:13 PM, Gabriel A. Devenyi <gdevenyi at gmail.com>
wrote:

> Thanks Andrew/Vlad,
>
> Vlad: that's a good question, I think what the apps are doing is
> computing the principal axes of the volume distribution, and aligning
> those with the slice directions
>
> Andrew: thanks, that's helpful if I ever run into that case.
> --
> Gabriel A. Devenyi B.Eng. Ph.D.
> Research Computing Associate
> Computational Brain Anatomy Laboratory
> Cerebral Imaging Center
> Douglas Mental Health University Institute
> Affiliate, Department of Psychiatry
> McGill University
> t: 514.761.6131x4781
> e: gdevenyi at gmail.com
>
>
> On Wed, May 4, 2016 at 1:44 PM, Andrew Wood <andrew at biospective.com>
> wrote:
> > Hi Gabriel,
> >
> > I think you need to resample the image such that the voxel lattice lines
> up
> > with the world dimensions. Would this do it?
> >
> > $ mincresample in.mnc -dircos 1 0 0  0 1 0  0 0 1 out.mnc
> >
> > - Andrew
> >
> >
> > On Wed, May 4, 2016 at 1:36 PM, Vladimir S. FONOV <
> vladimir.fonov at gmail.com>
> > wrote:
> >
> >> What is the definition of orthogonalized view?
> >>
> >> perhaps itk_resample --uniformize 1 will do it for you?
> >>
> >>
> >>
> >> On 16-05-04 01:22 PM, Gabriel A. Devenyi wrote:
> >>
> >>> Hi minc-users,
> >>>
> >>> Many nifti tools are "helpful" in that they orthogonalize your view
> >>> relative to the principal axes they compute from a native scan on
> >>> load. This makes it hard to compare to Display.
> >>>
> >>> Obviously, Display doesn't do this, but is there a tool or method to
> >>> compute such a transform so I can apply it to a MINC file?
> >>>
> >>> Thanks!
> >>>
> >>> --
> >>> Gabriel A. Devenyi B.Eng. Ph.D.
> >>> Research Computing Associate
> >>> Computational Brain Anatomy Laboratory
> >>> Cerebral Imaging Center
> >>> Douglas Mental Health University Institute
> >>> Affiliate, Department of Psychiatry
> >>> McGill University
> >>> t: 514.761.6131x4781
> >>> e: gdevenyi at gmail.com
> >>> _______________________________________________
> >>> MINC-users at bic.mni.mcgill.ca
> >>> http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-users
> >>>
> >>>
> >>
> >> --
> >> Best regards,
> >>
> >>  Vladimir S. FONOV ~ vladimir.fonov <at> gmail.com
> >> _______________________________________________
> >> MINC-users at bic.mni.mcgill.ca
> >> http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-users
> >>
> > _______________________________________________
> > MINC-users at bic.mni.mcgill.ca
> > http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-users
> _______________________________________________
> MINC-users at bic.mni.mcgill.ca
> http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-users
>