additional teaching file information template - DICOM

This is a discussion on additional teaching file information template - DICOM ; In the additional teaching file information template on p.39 of the following document: http://www.ihe.net/Technical_Framewo...2005-04-22.pdf For many of the attributes (e.g. keywords, finding, differential diagnosis/impression, etc.), there is both a text AND a coded value. Should there only be either a ...

+ Reply to Thread
Results 1 to 5 of 5

Thread: additional teaching file information template

  1. additional teaching file information template

    In the additional teaching file information template on p.39 of the
    following document:

    http://www.ihe.net/Technical_Framewo...2005-04-22.pdf

    For many of the attributes (e.g. keywords, finding, differential
    diagnosis/impression, etc.), there is both a text AND a coded value.
    Should there only be either a text OR a coded value and not BOTH?

    Thanks.
    Denny Lau


  2. Re: additional teaching file information template

    Every code in DICOM always requires both the code AND text. The text is
    more like a "courtesy", e.g. when you don't have the datadictionary
    available to interpret the code. Also, remember, text strings are
    notorious for being inconsise and are obvious in one particular
    language (e.g. text could be in french in certain parts of Canada)
    while the the code is international.

    Herman O.

    lau_denny@yahoo.com wrote:
    > In the additional teaching file information template on p.39 of the
    > following document:
    >
    > http://www.ihe.net/Technical_Framewo...2005-04-22.pdf
    >
    > For many of the attributes (e.g. keywords, finding, differential
    > diagnosis/impression, etc.), there is both a text AND a coded value.
    > Should there only be either a text OR a coded value and not BOTH?
    >
    > Thanks.
    > Denny Lau



  3. Re: additional teaching file information template

    Hi Herman

    I don't think that is the question that Denny was asking.

    In the template in the profile there are both CODE and TEXT
    content items for the same concept, and each row is specified
    as "1-n" and "U".

    That means there may be 0-n CODE and 0-n TEXT content items
    for each of the common coded concepts.

    However, one would not normally replicate the content, since a
    system rendering the content would then present duplicate information,
    and the text can always be extracted from the code (since as
    Herman points out the text part of the code is always required
    when the code is present)

    E.g., to say the Pathology is "Invasive lobular carcinoma" one
    could either send a TEXT content item with a value of "Invasive
    lobular carcinoma", or a CODE content item with a value of
    (SNM3, M-85203, "Invasive lobular carcinoma"), but not both.

    The intent is to allow the send to enter either codes or text,
    or a mixture (for things that are only partially coded).

    One could argue though that there is a use for saying the same
    thing twice but with multiple codes (that mean the same thing)
    from different coding systems ... e.g., if one wanted to send
    coded Keywords that were the same thing in ACR and MESH and ICD9CM
    (as opposed to "or"), and presumably in future, RadLex. That is
    not explicitly forbidden.

    David

    herman o. wrote:
    > Every code in DICOM always requires both the code AND text. The text is
    > more like a "courtesy", e.g. when you don't have the datadictionary
    > available to interpret the code. Also, remember, text strings are
    > notorious for being inconsise and are obvious in one particular
    > language (e.g. text could be in french in certain parts of Canada)
    > while the the code is international.
    >
    > Herman O.
    >
    > lau_denny@yahoo.com wrote:
    >> In the additional teaching file information template on p.39 of the
    >> following document:
    >>
    >> http://www.ihe.net/Technical_Framewo...2005-04-22.pdf
    >>
    >> For many of the attributes (e.g. keywords, finding, differential
    >> diagnosis/impression, etc.), there is both a text AND a coded value.
    >> Should there only be either a text OR a coded value and not BOTH?
    >>
    >> Thanks.
    >> Denny Lau

    >


  4. Re: additional teaching file information template

    Thank you for the clarification. My question was fully addressed.
    yes, I was confused about the VM and Req Type columns and which
    scenarios would require both TEXT and CODE values. Your examples
    clearly illustrated the implementation raionale.


    David Clunie wrote:
    > Hi Herman
    >
    > I don't think that is the question that Denny was asking.
    >
    > In the template in the profile there are both CODE and TEXT
    > content items for the same concept, and each row is specified
    > as "1-n" and "U".
    >
    > That means there may be 0-n CODE and 0-n TEXT content items
    > for each of the common coded concepts.
    >
    > However, one would not normally replicate the content, since a
    > system rendering the content would then present duplicate information,
    > and the text can always be extracted from the code (since as
    > Herman points out the text part of the code is always required
    > when the code is present)
    >
    > E.g., to say the Pathology is "Invasive lobular carcinoma" one
    > could either send a TEXT content item with a value of "Invasive
    > lobular carcinoma", or a CODE content item with a value of
    > (SNM3, M-85203, "Invasive lobular carcinoma"), but not both.
    >
    > The intent is to allow the send to enter either codes or text,
    > or a mixture (for things that are only partially coded).
    >
    > One could argue though that there is a use for saying the same
    > thing twice but with multiple codes (that mean the same thing)
    > from different coding systems ... e.g., if one wanted to send
    > coded Keywords that were the same thing in ACR and MESH and ICD9CM
    > (as opposed to "or"), and presumably in future, RadLex. That is
    > not explicitly forbidden.
    >
    > David
    >
    > herman o. wrote:
    > > Every code in DICOM always requires both the code AND text. The text is
    > > more like a "courtesy", e.g. when you don't have the datadictionary
    > > available to interpret the code. Also, remember, text strings are
    > > notorious for being inconsise and are obvious in one particular
    > > language (e.g. text could be in french in certain parts of Canada)
    > > while the the code is international.
    > >
    > > Herman O.
    > >
    > > lau_denny@yahoo.com wrote:
    > >> In the additional teaching file information template on p.39 of the
    > >> following document:
    > >>
    > >> http://www.ihe.net/Technical_Framewo...2005-04-22.pdf
    > >>
    > >> For many of the attributes (e.g. keywords, finding, differential
    > >> diagnosis/impression, etc.), there is both a text AND a coded value.
    > >> Should there only be either a text OR a coded value and not BOTH?
    > >>
    > >> Thanks.
    > >> Denny Lau

    > >



  5. Re: additional teaching file information template

    However, the standard DICOM templates for "saying the same thing twice
    with different coding systems" is to use TID 1210 Equivalent Meaning of
    Concept Name and TID 1211 Equivalent Meaning of Value. Both of these
    use a child node of the SR content tree (with a HAS CONCEPT MOD
    relationship to the primary encoding), rather than coding the alternate
    representation as a sibling node at the same level of the tree.

    - Harry

    David Clunie wrote:
    > One could argue though that there is a use for saying the same
    > thing twice but with multiple codes (that mean the same thing)
    > from different coding systems ... e.g., if one wanted to send
    > coded Keywords that were the same thing in ACR and MESH and ICD9CM
    > (as opposed to "or"), and presumably in future, RadLex. That is
    > not explicitly forbidden.



+ Reply to Thread