[MINC-users] Fwd: Minc to Analyze issue

[Ernest Lo]

Hi Andrew,

I had already edited the direction cosine information in the minc image (as
per a previous recommendation you made on minc-users):
minc_modify_header -dinsert xspace:direction_cosines=1,0,0 foo.mnc etc., so
hopefully that is not the problem.

With regard to the spm_defaults issue, I am using standard minc and analyze
formats -- is there a way to determine if spm_defaults is being used?

>>It could be the direction cosiones and it might also be that you are
>>using the spm_defaults type approach to handle left-right flips "radio
>>and neuro" orientation.  In which case, all bets are off.

I have been doing a lot of testing, and in one test case, minc2analyze
preserves image alignment across format conversion, whereas mnc2nii does
not.
I can show/send you the images if you like.  I'm still trying to figure out
where minc2analyze comes from in the context of minc tools.

>>minc2analyze?  If you mean mnc2ana || ana2mnc then they are now old
>>unsupported perl scripts of mine that may produce the same results and
>>may not dependant on where your "analyze" file came from.

The reason I would like to conserve image alignment across format conversion
(in both the images and transforms) is that I am experimenting with using
customized pre-processing (in minc format) for Siena (which requires analyze
format input).

Thanks a lot,

Ernest


Andrew Janke a.janke at gmail.com
Fri Jan 12 06:20:30 EST 2007

    * Previous message: [MINC-users] Fwd: Minc to Analyze issue
    * Next message: [MINC-users] xfm to mat ??
    * Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

> 1. It seems that the mnc2nii and nii2mnc are the preferred scripts to
> convert between minc and analyze formats.
> Can these scripts use and produce analyze (i.e. *.img and *.hdr) format
> files in addition to nifti (*.nii) format
> files?

Yes, use 'mnc2nii -help' to see the various output options.

> 2. We currently make use of the scripts minc2analyze and analyze2minc to
> perform conversions, I believe that are
> part of the minc tools.  How would these scripts compare with mnc2nii and
> nii2mnc?

minc2analyze?  If you mean mnc2ana || ana2mnc then they are now old
unsupported perl scripts of mine that may produce the same results and
may not dependant on where your "analyze" file came from.

> 3. I currently need to align two analyze format images.  I first convert
> them to minc format and then run an algorithm that produces the xfms.
Next
> I convert the xfm files to .mat format.  However the .mat transforms do
not
> align the original analyze format images.  I'm guessing that this is a
> result of the missing direction cosine information that pertains to
analyze
> format images.  However I would like to know if there is
> a way to convert from analyze to minc with no changes to alignment.

It could be the direction cosiones and it might also be that you are
using the spm_defaults type approach to handle left-right flips "radio
and neuro" orientation.  In which case, all bets are off.  My
suggestion would be that if you really do want to mess with both
analyze and minc would be to register and then resample in MINC before
converting back.


-- 
Andrew Janke   (a.janke at gmail.com || http://a.janke.googlepages.com/)
Canberra->Australia    +61 (402) 700 883