Internet Autopsy Database
Prototype Autopsy Diagnosis
UMLS Paragraph Translator.
All text is in the public domain,
and may be used freely without restriction.
Please select the desired sentence,
and click on the SUBMIT button:
Introduction to the UMLS Paragraph Translator.
1.
In autopsy pathology, the final results are typically
summarized as diagnostic PARAGRAPHS in a standarized format.
This summary form appears at the beginning of the
autopsy report, and is called the AUTOPSY FACESHEET.
Usually each medical institution or
private pathology practice has its own reporting form.
In the United States Government,
all pathology results are reported on
U. S. Standard Form 515,
an uncopyrighted, publicly available document.
2.
The most popular coding vocabulary
for pathology specimens is
SNOMED,
the Systematized Nomenclature
of Human and Veterinary Medicine
SNOMED is a copyrighted product of the
College of American Pathologists (CAP).
There are restrictions on the distribution
of SNOMED code numbers, which must be negotiated with CAP,
and typically require the payment of a user fee.
3.
The concepts within SNOMED have been contributed to the
Unified Medical Language System (UMLS) of the
United States National Library of Medicine (USNLM).
The UMLS is a metathesaurus,
containing over 600,000 concepts,
with over 2 million synonymous terms,
partial translations into German, French, Spanish,
Italian, Portuguese, Danish, Finnish, Swedish, Norwegian,
Dutch, Hungarian, and Romanized-Russian,
and linked to over 50 contributed biomedical terminologies.
In research investigations,
such as the Internet Autopsy Database,
the concept unique identifiers (CUIs)
of the UMLS may be discussed and
copied freely in public forums.
7.
The paragraph translator is
a public-domain computer translator,
with source code available in
PERL and
M, formerly MUMPS.
8.
In the translator, each word or multiple-word term
is pointed to its corresponding
UMLS concept.
In general, the longest consecutive sequence of words
is translated by the translator.
9.
For foreign language translation,
each concept may be mapped to a part-of-speech,
and the stereotypical word order
of the source language is rearranged
into the stereotypical word order
of the target language,
using a set of parsing formulas.
Since the source language (medical English)
and the target language (UMLS,
whose base language is English)
have the same stereotypical word order,
there is no word order rearrangement required.
Syntax Considerations for the UMLS Paragraph Translator.
1.
Within UMLS, there are atomic concepts,
e.g., "acute" (C0205178),
"inflammation" (C0021368),
and "appendix" (C0003617),
and compound concepts,
e.g., "acute appendicitis" (C0085693).
2.
In the IAD-UMLS lexicon, "acute appendicitis"
may be defined as C0085693
or alternatively as
C0205178!C0021368!C0003617,
using ! (ASCII 33) as a connecting character.
Cases are indexed redundantly as:
C0085693, C0205178, C0021368, C0003617.
3.
The IAD-UMLS translator contains additional
definitions not authorized by the USNLM, including:
Punctuation and linguistic particles
are classified as FULL STOP (C9999999),
as NULL (C0000000), or as COMMA (C9999998).
FULL STOP includes: . [period] : [colon]
; [semicolon] ! [exclamation point]
? [question mark]
( [left parenthesis] ) [right parenthesis] .
NULL includes: - the a these this that those I we you he she it they them.
In addition, medical concepts that we
think should be added to UMLS are designated as
EMERGENT (C9999997).
Concepts which have not yet received
a dictionary entry (an error on our part)
are designated as: CXXXXXXX .
Files in the IAD UMLS Paragraph Translator.
1. U. S. Government Tissue Examination Form
(U. S. Standard Form 515).
http://www.autopsydb.org/iadbs515.htm
2. Suite of PERL Translation Programs.
http://www.autopsydb.org/imarperl.htm
3. Suite of MUMPS Translation Programs.
http://www.autopsydb.org/imarmump.htm
4. List of Barrier Words.
http://www.autopsydb.org/imarbarr.htm
5. List of Keywords.
http://www.autopsydb.org/imarkeyw.htm
6. MRCON Subset.
http://www.autopsydb.org/imarmrco.htm
7. List of Parts-of-speech.
http://www.autopsydb.org/imarposp.htm
8. List of Medical Word Roots.
http://www.autopsydb.org/imarroot.htm
9. List of UMLS Pathology vocabulary.
http://www.autopsydb.org/umlslexi.htm
For further information,
please contact: G. William Moore, M.D., Ph.D.,
moore.md.george.w@baltimore.va.gov
Last updated: May 22, 1999.