Mif To Dicom


#1

Hi MRtrixler,

I was wondering, if there is the possibillity to change the mif files back to DICOM? I know, its difficult for some file-kinds. But maybe somebody needed it for FA or something and… by coincidence… you build it in???
I think it will be needed in the, I guess, far future! :slight_smile:

Greetings and sorry that I had no question sfor a long time, (because of your good work)

RALF


#3

Hey Ralf,

I can answer you directly with a figure from our website (somewhere):

…the devil is in the details: look at the arrowheads. :wink: (so, indeed: “nope” :frowning_face:)

Cheers,
Thijs


#4

To elaborate: this is not a simple matter though. @jdtournier can probably tell you great stories about the struggle to support as many possible manufacturers’ and scanner models’ DICOM formats. Going the other way around (i.e. to DICOM) would require to follow a certain convention in some way, but the problem is that there’s no single unique convention… on a lot of things that are very relevant to diffusion MRI data, e.g. how the gradient directions and b-values are stored, among others…


#5

Thanks Thijs,

thats what I thought. :cry:
So I have to go the way to wrap mif up and fake DICOM until ACR/NEMA-Standards are enhanced.

Thanks for the very quick answer,

Ralf


#6

Just to follow up on that: I used to have a command to replace the data in an existing DICOM image with that in another image with matching dimensions, but it’s fragile (doesn’t work for all DICOM files), and doesn’t handle any changes in image properties. Not something I’d generally advocate. Besides, I can’t find that code any more…

But otherwise, yes, DICOM is a nightmare, but the new standards are fairly settled now, so it’s something we could consider. But the output would be very basic, and there’s a good chance it won’t work on all systems. For instance, should we write out in enhanced / multiframe format, or stick to the old multi-file single-frame format? If someones’ keen on helping to do this, I’d be happy to assist. The bulk of the work is to figure out which DICOM tags are mandatory and need to be written out, and what values they should take, etc. Not altogether very difficult, but pretty tedious nonetheless…

If you’re keen, get in touch!