Dear Daniel,
I want to thank you for taking the time to check the images. Yes, the
images have some processing (som edge delineation), but I also have the
original ones and it happens the same to them; actually, the culprits here,
the ImageOrientationPatient (the direction cosines) and the
ImagePositionPatient fields are the same (I've also been tracing the
DicomSeriesReader code and I've found the place that you're mentioning,
where it detects the shearing and loads the slices as separate images).
This kind of shearing is really common with the MRI images we usually see.
Actually, these volumes are correctly loaded by ITK-Snap and some
commercial applications. Also, if I use the DICOM editor plugin, it loads
this same DICOM as a volume. It shows it a little bit skewed in the
sagittal and axial planes, because it is not correcting the shearing, but
after doing a reinit it corrects it and shows the volume exactly as
ITK-Snap and the commercial applications do.
I do not know what reinit is exactly doing, I've not followed it, but I
think that if we could apply the same transform that reinit calculates to
the volume, we can correct the shearing when we're loading it.
Best,
Federico
On Tue, Mar 26, 2013 at 1:24 PM, Daniel Maleike
<[email protected]>wrote:
> On 11.03.2013 15:50, Federico Milano wrote:
> > Hi. I'm using the ProjectTemplate application to load a MRI DICOM series
> > representing a volume in the Data Manager. The problem is that each
> > image is loaded as a different Data Node, and not as a single volumetric
> > image (this also happens in the workbench application). On the other
> > hand, if I load it in the MITK 3M3 Data Manager, the series is
> > recognized as a volume and loaded as a single data node. I know that
> > MITK 3M3 is closed source and developed by a private company, but do you
> > have any clue if there is some difference in the way DICOM files are
> > treated? I'm working in research and I would really need to load these
> > images as a volume.
>
> Dear Federico,
>
> I just had a look at the images you provided. Thanks for loading them up.
>
> Other than "2D DICOM viewers", MITK always interprets DICOM images as
> images in 3D space. The loader tries to group slices by similar tags and
> sorts them spatially, taking into account the tags
>
> 0020,0032 ImagePositionPatient
> 0020,0037 ImageOrientationPatient
>
> I tried plotting the positions and orientations of your images and they
> look something like (ascii art)
>
> ------------------------
> ------------------------
> ------------------------
> ------------------------
> ------------------------
>
> From what I understood, the shift parallel to your plane orientation is
> happening in two dimensions, so we don't have a situation as in gantry
> tilt images.
>
> MITK attempts quite some sorting but ends up with loading each image
> individually, because we cannot group them into 3D volumes.
>
> Two questions:
> - how did you get these images? They look processed. Any chance some
> step in your pipeline modified plane positions?
> - did you succeed to load these images "as a volume" with any other
> kind of software?
>
> Probably a better loader for your images could be implemented in MITK,
> but currently there is no effort to do so. If you want to add to such a
> solution, I recommend reading mitkDicomSeriesReader.cpp (the tilt part)
> and working on the tilt detection/correction code.
>
> Kind regards
> Daniel
>
>
> --
> Dr. Daniel Maleike Telefon: + 49 6221 647976 3
> Mint Medical GmbH, Friedrich-Ebert-Straße 2, 69221 Dossenheim/Heidelberg
> Geschäftsführer: Dr. Matthias Baumhauer
> Registergericht Mannheim, HRB 709351
>
>
> ------------------------------------------------------------------------------
> Own the Future-Intel® Level Up Game Demo Contest 2013
> Rise to greatness in Intel's independent game demo contest.
> Compete for recognition, cash, and the chance to get your game
> on Steam. $5K grand prize plus 10 genre and skill prizes.
> Submit your demo by 6/6/13. http://p.sf.net/sfu/intel_levelupd2d
> _______________________________________________
> mitk-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/mitk-users
>
------------------------------------------------------------------------------
Own the Future-Intel® Level Up Game Demo Contest 2013
Rise to greatness in Intel's independent game demo contest.
Compete for recognition, cash, and the chance to get your game
on Steam. $5K grand prize plus 10 genre and skill prizes.
Submit your demo by 6/6/13. http://p.sf.net/sfu/intel_levelupd2d
_______________________________________________
mitk-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mitk-users
