[MINC-users] conversion to analyze problem

Sat Sep 8 10:34:04 EDT 2007

Hi Andrew,

Although I am sympathetic to systemic problems and possible
ambiguities of the MINC <-> ANALYZE format conversions, I am as yet
not convinced that we are not looking at a bug; nor that the behaviour
of mnc2nii/nii2mnc can/should not be improved to "work" for users in
95% of the cases instead of 5% (if that). Reasons being:

1. I am yet to encounter the first case where flipping was not
necessary (and I have done my fair share - probably more - of these
types of conversions)

2. If I take a MINC volume (without funky directions cosines and etc),
run mnc2nii (-analyze) followed by nii2mnc, the x-axis comes out
flipped.

3. In the case that Lindsay reports, at least 3 different analyze
viewers agree that the x-axis is flipped (which I understand not to be
an issue of viewer/radiological/neurological convention, as evidenced
by a physical marker in the image).

So notwithstanding MINC's conceptual superiority w/respect to
coordinate space handling and etc, it still makes no sense to me that
this conversion appears to not *ever* work correctly, nor that the
conversion on 'simple' files is not bijective when using the
conversion tools provided by MINC itself. Something just smells wrong
here...

Where does one actually find the formal specifications of
ANALYZE/NIFTi? For ANALYZE I keep stumbling on the same pdf describing
the ANALYZE 7.5 file format, but that is less than informative.

-- A

On 9/7/07, Andrew Janke <a.janke at gmail.com> wrote:
> Hi Lindsay, (and others!)
>
> As Alex has alluded to, the problems with "analyze" to MINC conversion
> are systemic and not really bugs.
>
> In short there is no way to canonically convert between the two
> formats and ensure that the orientation will always be correct as it
> depends on the package that you will be using with your
> "analyze/Nifti" images. What generally happens is that users have to
> figure out a path that works for them and stick to it.
>
> This in itself is endemic in any Nifti/Analyze conversion but what I
> can assure you of are the following...
>
> 1) mnc2nii follows the official specifications of Analyze and Nifti.
>
> 2) The issue of Radiological vs Neurological orientation _should_ have
> nothing to do with a file format and only concern a file viewer. In
> MINC this is the case. There is also a defined co-ordinate system for
> ANALYZE and Nifti-1 but these are not always followed to the letter.
>
> 3) Further to #2 you need to very careful of what you package is doing
> to your files as some like SPM define a file like "spm_defaults.m"
> that defines as to whether your data is "neurological" or
> "radiological" that will flip your data inside the package itself.
>
> Sorry I cant offer a simple answer to your problems but the above is
> one of the primary motivating factors that caused Peter (Neelin) to
> write MINC nigh on 12 years ago! :)
>
> That all said there is a new MINC due out very soon that will have an
> updated version of Nifti conversion that may alleviate your concerns
> as we  now use the same version of the Nifti Libraries that is used by
> FSL and thus supports a few extra things like compression of Nifti
> files and the likes.
>
>
> a
>
>
> On 9/8/07, Walker, Lindsay (NIH/NICHD) [V] <walkerlin at mail.nih.gov> wrote:
> > That definitely helps.  I will give it a try.
> >
> > Is there a newer version of this tool where this bug has been addressed?
> > Or is it possible for someone to fix this bug?
> >
> > Lindsay.
> >
> > -----Original Message-----
> > From: Najmeh Khalili M. [mailto:najma at bic.mni.mcgill.ca]
> > Sent: Friday, September 07, 2007 3:49 PM
> > To: MINC users mailing list
> > Subject: Re: [MINC-users] conversion to analyze problem
> >
> > Hi,
> >
> > Yes I have too.
> > I did something like:
> >              fslswapdim data.nii.gz x -y z flipped_data.nii.gz
> > to correct the left-right flip problem.
> >
> > On Fri, 7 Sep 2007, Walker, Lindsay (NIH/NICHD) [V] wrote:
> >
> > > I need to convert some minc images to analyze format.  I have been
> > using
> > > the mnc2nii tool from minc-1.5.1 with the -analyze option.
> > >
> > > The images I converted have an indicator dot on the anatomical right
> > of
> > > the image.  When I run the minc to analyze conversion program, the
> > > resulting analyze image shows the indicator dot on the anatomical
> > left.
> > > So the program appears to be flipping the labels for the left and
> > right.
> > >
> > > I have tried opening the analyze image with multiple pieces of
> > software
> > > to make sure it is not just a display issue (i.e. radiological versus
> > > neurological coordinates).  The left and right labels appear to be
> > > incorrect using fslview, medx and mipav.
> > >
> > > Has anyone else encountered or noticed this problem?
>
>
> --
> Andrew Janke   (a.janke at gmail.com || http://a.janke.googlepages.com/)
> Canberra->Australia    +61 (402) 700 883
> _______________________________________________
> MINC-users at bic.mni.mcgill.ca
> http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-users
>
>