Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Equivalent meaning of concept name AND Language of Content Item and Descendants

32 views
Skip to first unread message

Bartosz Wiklak

unread,
Jun 30, 2009, 6:58:02 AM6/30/09
to
Hi,
I made simple SR, dcmtk dsrdump shows this output for it:

Basic Text SR Document

Patient : MATEJKO^JAN (M, 18380624, #1239)
Study : Opis badania
Completion Flag : PARTIAL
Verification Flag : UNVERIFIED
Content Date/Time : 20090630 124504

<CONTAINER:(,,"Radiology Report")=SEPARATE>
<has concept mod TEXT:(,,"Equivalent meaning of concept name")="Opis
badania radiologicznego">
<has concept mod CODE:(,,"Language of Value")=
(pol,ISO639_2,"Polish")>
<has concept mod CODE:(,,"Language of Content Item and Descendants")=
(pol,ISO639_2,"Polish")>
<has obs context CODE:(,,"Observer Type")=(121006,DCM,"Person")>
<has obs context PNAME:(,,"Person Observer Name")="Test^Testowy">
<has concept mod TEXT:(,,"Equivalent meaning of concept name")
="Imię i nazwisko osoby opisu...">
<has concept mod CODE:(,,"Language of Value")=
(pol,ISO639_2,"Polish")>
<contains CONTAINER:(,,"Summary")=SEPARATE>
<has concept mod TEXT:(,,"Equivalent meaning of concept name")
="Podsumowanie">
<has concept mod CODE:(,,"Language of Value")=
(pol,ISO639_2,"Polish")>
<contains TEXT:(,,"Summary")="To jest text">
<has concept mod TEXT:(,,"Equivalent meaning of concept name")
="Podsumowanie">
<has concept mod CODE:(,,"Language of Value")=
(pol,ISO639_2,"Polish")>


I have two questions:
1. Does this example conforms IHE SINR (does it fallow DTID 2000)?
2. How should I provide national concept name meanings - should I do
it like in example above or maybe I should set "Language of Content
Item and Descendants" to English and use "Equivalent meaning of
concept name" like above or else, I should set "Language of Content
Item and Descendants" to Polish like in example but not use
"Equivalent meaning of concept name" rather than use national names in
code meanings?

I really tend to do things "the right way", but paths of dicom are not
strait, not at all :(
Bartek

David Clunie

unread,
Jul 5, 2009, 10:33:50 AM7/5/09
to
Hi

What makes this complicated in the general case is deciding on one of
two strategies:

- everything below and including a content item are in the same
language, including BOTH concept name and value (TID 1204), or

- some content items have a different language for either the value
(only) (TID 1201) or both the concept name and value (TID 1202)

A variant is whether or not to include the same information in
multiple languages, in which case you also have a choice, for each
content item, of:

- equivalent meaning of concept name (TID 1210)

- equivalent meaning of value(s) (TID 1211)

Now, some root level templates (e.g., IHE SINR aka. TID 2000)
constrain your choices here; specifically, these templates are non-
extensible (you cannot add your own additional content items) in order
to simplify the patterns that the recipient has to deal with for
rendering. In this case, TID 2000 allows for the document root
container (report title) ONLY:

- TID 1204, language of content item and descendants

- TID 1210 Equivalent Meaning of Concept Name

but does NOT allow for further qualification of anything below the
root (i.e., one cannot provide equivalent meanings or selectively
different languages for concept names or values below the root, e.g.,
for section headings).

So, for your example, if you want to be IHE SINR compliant, and you
want to the report payload to be in Polish, then you need to also have
the report title, and ALL of the concept names (not just values) in
Polish as well. You could also provide an equivalent meaning of the
report title concept name in English, if you wanted to, but there
would be little point in this, since the code is what matters for
recognition (the meaning is only a rendering convenience in the
absence of a local dictionary).

So, using your syntax, this might look like:

<CONTAINER:(LN,11528-7,"Radiologia Raport")=SEPARATE>
<has concept mod CODE:(121049,DCM,"Język Treść pkt i Descendants")=
(pol,RFC3066,"Polski")>
<has concept mod TEXT:(121050,DCM,"Równoważne znaczenie pojęcia
nazwa")="Radiology Report">
<has concept mod CODE:(121047,DCM,”Język wartości”)=
(eng,RFC3066,"English")>
<has obs context CODE:(121005,DCM,"Obserwator Typ")=
(121006,DCM,"Osoba")>
<has obs context PNAME:(121008,DCM,"Nazwisko osoby obserwatora")
="Test^Testowy">
<contains CONTAINER:(DCM,121111,"Podsumowanie")=SEPARATE>
<contains TEXT:(DCM,121111,"Podsumowanie")="To jest text">
...

Please excuse my crude attempts at Polish translation.

The practical consequence of this is that SINRs generated in the local
language are easily rendered for the local users, BUT it requires the
creator to translate ALL of the concept names used into the local
language if DICOM hasn't already done so.

If you do this for Polish, then send the translations to me, and I
will add them to the tables in PS 3.16 for other folks to use.

David

Note also the use of RFC3066 rather than ISO639_2 for the language
code (see CP 371).

David

On Jun 30, 6:58 am, Bartosz Wiklak <bwik...@gmail.com> wrote:
> Hi,
> I made simple SR, dcmtk dsrdump shows this output for it:
>
> Basic Text SR Document
>
> Patient             : MATEJKO^JAN (M, 18380624, #1239)
> Study               : Opis badania
> Completion Flag     : PARTIAL
> Verification Flag   : UNVERIFIED
> Content Date/Time   : 20090630 124504
>
> <CONTAINER:(,,"Radiology Report")=SEPARATE>
>   <has concept mod TEXT:(,,"Equivalent meaning of concept name")="Opis
> badania radiologicznego">
>     <has concept mod CODE:(,,"Language of Value")=
> (pol,ISO639_2,"Polish")>
>   <has concept mod CODE:(,,"Language of Content Item and Descendants")=
> (pol,ISO639_2,"Polish")>
>   <has obs context CODE:(,,"Observer Type")=(121006,DCM,"Person")>
>   <has obs context PNAME:(,,"Person Observer Name")="Test^Testowy">
>     <has concept mod TEXT:(,,"Equivalent meaning of concept name")

> ="Imiê i nazwisko osoby opisu...">

0 new messages