MUGIT - Medical University of Graz: Interface terminology for SNOMED CT
http://user.medunigraz.at/stefan.schulz/mugit/

This directory contains filtered versions of the experimental German Interface Terminology linked to the SNOMED CT version in several releases (see directory "releases"), and the current, frequently changing version (see directory "current").
(c) Medical University of Graz
Contact: Stefan Schulz - stefan.schulz@medunigraz.at

- Filtered Interface terminology versions
  Terminology, filtered according to frequency and existence of terms and term substrings in German 
  reference corpora and terminologies
  - SCT-GIT_de_precise.dat: strict term matching criteria 
   (max 20 terms per concept, minimal score 0.9, no length restrictions)
  - SCT-GIT_de_large.dat: more relaxed term matching criteria,
    contains many nonsensical terms 
    (max 100 terms per concept, minimal score 0.55, no length restrictions)
  - SCT-GIT_de_nlp.dat: relaxed term matching criteria: optimised for used for text mining 
    (max 100 terms per concept, minimal score 0.55, maximum 5 tokens / term)
  - SCT-GIT_de_drugs_nlp.obo: relaxed term matching criteria: optimised for used for text mining 
    within Averbis Health Discovery, using drug information from the official Austrian source
    Arzneispezialitätenregister (ASP)   

- Release policy, naming
  - the current file has always the above names. The files of the most recent release are in 
    the releases subdirectory, previous ones as zip archives with the respective time stamps
  - in the OBO file the creation time is in the header
  - the content in the "current" directory reflect the current state of the terminology and is 
    subject of frequent changes

Conditions of use:

This resource is still experimental and is provided "as is". It is frequently but irregularly updated. Until further notice, it can be used by third parties for research only, but not for operational applications without charges, usage fees, or royalties. The use is at the sole risk of the user and at no expense to the owner.

The owner reserves the right to disseminate a future, quality-assured version through a corporate distribution partner under license conditions to be defined. In this case, the experimental version will be discontinued. 

Whoever uses this resource has to communicate its use informally to the owner by E-mail. The owner is maintaining a registry of users.  A formal licence agreement is not required until further notice. The owner expects from the user regular feedback that helps improve the product, particularly the report of errors.  

- Related work

  MUGIT_MEDINFO_2019.pdf - Presentation at MEDINFO 2019 
  Hashemian Nik D, Kasac Z, Goda Z, Semlitsch A, Schulz S.
  Building an Experimental German User Interface Terminology Linked to SNOMED CT.
  Stud Health Technol Inform. 2019 Aug 21;264:153-157. doi: 10.3233/SHTI190202.
  http://ebooks.iospress.nl/publication/51967
  
  Schulz, S
  Annotating German Clinical Narratives with SNOMED CT
  https://confluence.ihtsdotools.org/display/FT/SNOMED+CT+Expo+2020. 2020; 
  SNOMED EXPO 2020; 08. - 09.10.2020; London, UK

  Schulz, S; Hammer, L; Hashemian-Nik, D; Kreuzthaler, M
  Localising the Clinical Terminology SNOMED CT by 
  Semi-automated Creation of a German Interface Vocabulary
  In: Melero, M editors(s). Proceedings of the LREC 2020 
  Workshop on Multilingual Biomedical Text Processing (MultilingualBIO 2020). 
  Luxemburg: European Language Resources Association; p. 15-20. 
  2020(ISBN: 979-10-95546-65-8)
  https://www.aclweb.org/anthology/2020.multilingualbio-1.3/

Format of .dat files:
Fields delineated by Tabs
UTF-8

Field #1: SNOMED Concept ID
Field #2: Internal Term ID suffixed by creation date in yyyymmdd format
Field #3: English Term
Field #4: German Term

Format of .obo files: see documentation of Averbis Health Discovery

Additional files: 
- Transitive closure file: each active SNOMED CT concepts with its ancestors: 
  annotation expansion for hierarchical search
- File with negative pairs (code, term, optionally FSN), tab delimited: 
  can be extended to supress codes for specific terms
- List of active FSNs
- Example code in Python (requires these files, produces a simple HTML output)
  for use with Averbis Health Discovery version 6

Last update: 2022-09-13 by Stefan Schulz