Page tree

Date & Time

20:00 UTC Wednesday 13th September 2017

Teleconference Details

To join the meeting please go to

Further information can be found at SLPG meeting information


  • Discuss implications of potential DL enhancements and other outstanding questions
  • Review proposed Query language examples

Agenda and Meeting Notes


Welcome and apologies

Questions from DanielAll

Discuss questions from Daniel

  1. to support expression repositories (i.e. where the substrate is dynamic in that expressions could be added), do we need to add the ability to use the ECL to apply constraint operators to post-coordinated expressions, i.e. where a piece of the ECL string is to be interpreted as a SCG expression, e.g. << [ 404684003 | Clinical finding (finding) |: 363698007 |Finding site (attribute)| = 10200004 | Liver structure (body structure) | ]? This would correspond to a DL query.
  2. to support potential expansion of the SNOMED CT logic profile, do we need to extend SCG to support axioms as opposed to only expressions, for example to allow (non-hidden) GCAs to be stated in SCG?
  3. we need an SNOMED CT OWL specification, i.e. how is SNOMED CT (concepts and descriptions) represented in OWL. Is this within the scope of this group?
Implications of enhanced DLAllDiscuss any implications to computable languages of potential enhancements to DL - for example:
  • GCIs - Clarification required as to whether non-necessary conditions will be in scope
    • For example, a concept |Disorder of the arm or leg| defined with 2 sufficient (but individually not necessary) conditions such as |Finding site|= |Arm structure|, |Finding site| = |Leg structure|
  • Role chains - Not an issue to ECL
  • Universal restriction - out of scope
  • Concepts as numbers - needs discussion (Do we need to transform these to concrete domains to compute?)
Query Language
  • Review proposed Query Language examples
  • Additional requirements from SLPG members?
Confirm next meeting date/time

Next meeting to be held at 20:00 UTC on Wednesday 8th November 2017 (due to Linda's travel commitments)

  • No labels


  1. Use case for item 2.3: Some of the LOINC-terms include disjunctions and could be represented using GCIs. This could be represented by multiple rows in the refset with the same mapTarget, but preferably by keeping the full definition in one row using multiple expressions. The left-most left-hand+definition-status would be represented separately in the refset outside of the expression. (Why was definitionStatusId given a special field when SCG can represent this?)

  2. Please note that Peris Brodsky has just published a new ECL implementation called 'Slang' -

    I have added this link to the SLPG ECL page. If you have any feedback, I'm sure he would appreciate your comments.

    Kind regards,

    1. Apologies! I spoke too soon. Slang will be up in the next couple of days ... so stay tuned! (smile)

    2. Slang is back in action at Please have a look and let Peris know if you have any feedback.