[Neuroimaging] JSON-LD and DICOM?

[Bennet Fauber]

Oh, yes.  That is very important.

Converting to .nii is one of our main ways of insuring no leakage of
any subject information -- not just PHI, but potentially names and
other information about research subject that, while not regulated,
would discourteous to expose.




On Tue, Jul 4, 2017 at 1:05 PM, David Gutman <dagutman at gmail.com> wrote:
> Just to chime in briefly.. one thing that wasn't mentioned re: DICOM is the
> fact of potential PHI leaks.    Ensuring any and all DICOM tags have no PHI
> in it is an incredibly tedious/expensive undertaking, and at least for those
> of us who use clinical scans for some of our work, the overhead of trying to
> make sure we truly anonymized everything in the DICOM header is significant.
>
> Whereas in the case of NIFTI, as long as the filename doesn't contain the
> embedded patient name , I can be pretty sure the files are "clean".
>
> This is particularly important when grad students are working on their own
> laptops which are... not particularly secure.
>
>
>
>
> On Tue, Jul 4, 2017 at 1:01 PM Chris Gorgolewski
> <krzysztof.gorgolewski at gmail.com> wrote:
>>
>> I wonder if in the future it would also make sense to:
>>
>> 1) add option of creating sidecar BIDS compatible JSON files into nibabel
>> (similar to what dcm2niix, dicm2nii, and mcverter are doing). This makes
>> some sense since nibabel will become a dicom to nifti converter after
>> dcmstack is merged.
>>
>> 2) store the same content in the header (for "future" compatibility) under
>> a different namespace than "dcm" (would that work Satra? How would an
>> example JSON-LD with both "dcm" and "bids" look like?)
>>
>> Not a huge priority, but something to think about. I am very glad that
>> dcmstack is getting merged into nibabel.
>>
>> Best,
>> Chris
>>
>> On Tue, Jul 4, 2017 at 12:47 PM, Satrajit Ghosh <satra at mit.edu> wrote:
>>>>
>>>> Excellent - that would be very helpful.  Will y'all be able to define
>>>> the @types as well as the @ids ?  I guess that would be a lot more
>>>> work.
>>>
>>>
>>> it shouldn't be too hard to combine it with this:
>>>
>>> https://github.com/pydicom/pydicom/blob/master/pydicom/_dicom_dict.py
>>>
>>> which itself is generated from the dicom standard and private
>>> dictionaries (these are not in karl's curation though and likely varies by
>>> sequence and scanner software versions).
>>>
>>> https://github.com/pydicom/pydicom/tree/master/source/generate_dict
>>>
>>> cheers,
>>>
>>> satra
>>>
>>> _______________________________________________
>>> Neuroimaging mailing list
>>> Neuroimaging at python.org
>>> https://mail.python.org/mailman/listinfo/neuroimaging
>>>
>>
>> _______________________________________________
>> Neuroimaging mailing list
>> Neuroimaging at python.org
>> https://mail.python.org/mailman/listinfo/neuroimaging
>
> --
> David A Gutman MD PhD
> Assistant Professor of Neurology, Psychiatry & Biomedical Informatics
> Emory University School of Medicine
> Staff Physician, Mental Health Service Line
> Atlanta VA Medical Center
>
> _______________________________________________
> Neuroimaging mailing list
> Neuroimaging at python.org
> https://mail.python.org/mailman/listinfo/neuroimaging
>