[Neuroimaging] JSON-LD and DICOM?

[Matthew Brett]

On Mon, Jul 3, 2017 at 1:00 PM, valabregue <romain.valabregue at upmc.fr> wrote:
>
>
> On 07/03/2017 11:54 AM, Matthew Brett wrote:
>>
>> Hi,
>>
>> On Mon, Jul 3, 2017 at 9:21 AM, valabregue <romain.valabregue at upmc.fr>
>> wrote:
>>>
>>> Hi all,
>>>
>>>
>>>
>>> On 06/30/2017 09:57 PM, Matthew Brett wrote:
>>>>
>>>> Hi,
>>>>
>>>> On Fri, Jun 30, 2017 at 5:35 PM, Jasper van den Bosch <japsai at gmail.com>
>>>> wrote:
>>>>>
>>>>> Hi all,
>>>>> I have to agree with Andrey on the yet another format argument. Also,
>>>>> tools
>>>>> will have to convert it to other formats anyway, so if you do end up
>>>>> storing
>>>>> something in the header, as long as you document it, just like you
>>>>> would
>>>>> other nibabel properties, I'd go for the simplest solution.
>>>>
>>>> There is going to be a JSON header extension quite soon, so it's not
>>>> really another format, but another way of storing metadata in NIfTI.
>>>> Then the question is - should we also store DICOM metadata there?
>>>
>>> Yes we should store the Dicom metadata, this is the less effort, and it
>>> fulfills the needs ... May be I miss the point. What would be the
>>> alternative ? (to define a new ontologie for the metadata ?)
>>>
>>> I am using (since a couple a years) the dcmstack that convert dicom to
>>> nifti
>>> + a json file (with all dicom meta data AND more importantly all private
>>> siemens fields). The nice feature of this conversion is that it
>>> aggregates
>>> field that are identical in all split dicom file (and keep array only
>>> when
>>> it is necessary).
>>>
>>> Why not use this solution ?
>>
>> Actually, the reason I'm interested in this now is because I am
>> thinking about how to integrate dcmstack into nibabel.  So I wanted to
>> fuse the dcmstack metadata analysis into the NIfTI header.
>>
>>> I do not see any clear advantage to have this meta data information
>>> directly
>>> in the nifti file (a naming convention make it easy to keep a link)
>>
>> Well - let's say there are two options:
>>
>> * A NIfTI file with a JSON header, where the JSON header contains the
>> DICOM metadata;
>> * A NIfTI file with a JSON header, where the the DICOM metadata is in
>> a separate JSON file.
>>
>> To me the first seems more obvious than the second.  Surely it also
>> makes things like provenance just a little bit easier - because it is
>> just a little bit harder to lose the relationship between the image
>> and the metadata.
>
> Well may be it is not a big deal to make both possible ... ?
>
> It looks like nifti files extension  img+hdf or .nii
> well I agree with you: all in once is more obvious. ( like for nifti file I
> prefer the .nii files). The only disadvantage I see is that you may loose
> some nifti compatibility with other software that expect only a fix header.
> (345 byte). So depending on what you plan to do, both may be chosen.

I think most if not all software allows there to be header extensions
after the header and before the data, so I think that will be OK...

Cheers,

Matthew