Page tree

Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Date & Time

Thursday 28th November 2024

Location

Zoom meeting link

Goals

  • Expand the use of the ECL language
  • Make ECL expansion consistent on well known FHIR Terminology Servers
  • Consider requirements for the maintenance of ECL and postcoordinated expressions

Agenda and Meeting Notes

Advanced Tables - Table Plus
border2
rowStylesbackground-color:#ccccff;font-weight:bold;,background-color:"#eeeeff";font-weight:normal;,background-color:#eeffff;font-weight:normal;
enableSortingfalse


Description

Owner

Notes

Welcome and agenda

All


ECL 2.2
Implementation Progress

  • "Top" and "Bottom" operators
    • Ontoserver has and Snow Owl have implemented this
    • Scheduled for development in Snowstorm
  • Alternate Identifiers
    • No implementations yet
    • The LOINC SNOMED Extension should be published in ~March 2025

ECL 2.3 development

Behaviour refinement for concrete string matching

Proposal: Simplify concrete string matching. Use exact string match rather than prefix matching.

Justification: The specification is a little vague in this area. The project group agreed that exact string matching seems most appropriate for concrete strings. The known terminology servers have already implemented exact string matching. 

- Group already reviewed documentation changes -

ECL 2.3 development

Ensuring consistency


Proposal. "exact:" string matching prefix.

TODO: Is this needed?

TODO: Should wildcard search be case sensitive? Can we document the behaviour.

Current term matching options:

  • {{ D term = "asth" }} equivalent to {{ D term = match:"diab mel" }}
  • {{ D term = wild:"as*th" }}
  • Should we add wild - {{ D term = exact:"asthma" }}

Compound word search



ECL 2.3 development

Ensuring consistency

All

Updates following Jeremy Roger's Terminology Terver ECL consistency analysis the following changes were recommended by the group:

  • Should FSN be excluded by default from Description filters?
    • Pros:
      • A lot of results when search term includes semantic tag word, e.g. disorder
        • Workaround: engines could use ranking to reduce the score of concepts matching by semantic tag
    • Cons:
      • Not all concepts have a synonym exactly matching the FSN minus its semantic tag .. .but the huge majority do. Only 14.6% do not, but even most of these have at least one vanilla synonym that is incredibly similar to the FSN (at least, for the language=en presentation of SNOMED). For this reason, adding the words of "FSN minus semantic tag" into the lexical search space by default almost never materially alters the lexical retrieval result, because there are enough words in the vanilla Synonyms. But adding the semantic tag itself in as well is at best redundant and at worst a source of false positive search results. If the rules and therefore benefits of including the FSN are different for searches over other languages then fine...but if that is the case then we must consider whether imposing language-blind rules and defaults on Terminology Server behaviours is a good idea. It appears to be adversely impacting anglophone search.

        Consider Rogers' Conjecture:
        If semantic tags had not been encoded within SNOMED CT in the way that they are, embedded within the string of FSNs, but had instead been encoded as either a standalone description type, or by the membership of a crefset, then we wouldn’t be dreaming of including them by default within the scope of lexical searches.

        If the words in the rest of the FSN, without the semantic tag, are presented in other synonyms anyway then, outside of very specialist terminologist use cases, there is virtually no reason to ever perform a lexical search on the FSN whether by default or for fun. If you want to constrain a search both semantically and lexically, use the taxonomy not the semantic tags for the semantic constraint element. If you just want to include the words in the FSN other than the semantic tag in your lexical search, then you've already got them in the synonyms. And if you haven't then there is probably a modelling/naming error that would be better fixed in the source data rather than compensated for post hoc by always including the FSN in searches even where it isn't needed and does more harm than good.

All

Updates following Jeremy Roger's Terminology Terver ECL consistency analysis the following changes were recommended by the group:



ECL Lite - Expanding the use of the ECL standard

All

Previous discussion:

How to make it easier to add an ECL capability to more tooling?

  • SI already has open source implementations for reference
  • The specification is now very large.
    • An ECL Lite profile would reduce the months of effort required to implement the whole specification when starting fresh.


"ECL Lite" is a simpler version of Expression Constraint Language including only the most useful features. ECL Lite will be a true subset of ECL so will be forward compatible. 


Feedback from previous meeting:

  • Concrete string matching simplification (resolved above)

  • Not including description-filters limits the ability to define a set of concepts by language and dialect
    • Consider publishing a recommendation to workaround this in FHIR Terminology servers.
    • Review: Added scope info box: 6.8 Description Filters

  • Add "Included In ECL Lite" Labels throughout the examples section - Use positive labels rather than exclusions
    • Review: Added throughout - 

  • Appendix - make explicit that ECL Lite syntax must work with servers with full - intention is true subset.

Other draft specification updates:



ECL Lite - MVP Scope:

  • *, <, <<, >, >>, <!, !>, :, ^, AND, OR, MINUS, (), {{ +HISTORY-(MIN,MOD,MAX) }}, <!!, !!>, {{ C active }}
  • Dot
  • Not included:
    • other filters
    • cardinality
    • groups
  • Declare profile using block/set of language capabilities
  • Survey capabilities in existing implementations - any existing capability groups?
  • There are parallels in ECL builder and education scopes


ECL Capability statement in FHIR


Chat with Peter G. Williams 

ECL Test Harness


Kai Kewley Márk Czotter 

Next steps:

FHIR TX - Searching with a specific dialect


Kai Kewley  Try this in Snowstorm:

When expanding a ValueSet with text filtering it is recommended to use the "Content-Language" header, to constrain the language and dialect of matched terms, in addition to informing the selection of display terms.

This behaviour is owned by HL7.

ECL Lite naming


Not everyone happy with the name.

TODO: Choose from alternatives:

  • ECL Core
  • ECL Essential
  • ECL Focus
  • ECL Base
  • ECL Basic
  • ECL--
  • My First ECL
Substrate discussion


Add appendix to describe the potential differences between terminology servers given what content is loaded/filtered out.

ECL Enhancement Request: Set of reference sets containing a concept

Enhancement for ECL 2.3

See comment here: Re: Discussions (2)

Use cases:

  • Exploration for data analytics
    • What subsets ~ what domains - how is this code used
  • System design - finding existing relevant sets
    • Would require a set of concepts that are all included within a refset
  • Browsing
    • Like the SI Browser "Refsets" tab

Syntax options:

  • ^[refsetId] * {{ M referencedComponentId = 404684003 }}
  • Syntax sugar: ^? 404684003

  • ^[refsetId] < 123000 {{ M referencedComponentId = 404684003 }}
  • Syntax sugar: ^? 404684003 AND < 123000

  • ^[refsetId] * {{ M referencedComponentId = << 404684003 }}
  • Syntax sugar: ^? << 404684003

Agreed. Kai Kewley draft ECL guide changes for review next time (Not done yet)

  • @404684003
    • Overlaps with template language
  • ? 404684003
    • A bit vague? 
  • ^? 404684003
    • Looks intuitive
    • Two Behaviours
    • refsets returned match ANY concepts in the nested expression.
      • ^? (<404684003)
      • ^? (404684003 OR 234234234002)
    • refsets returned match ALL concepts in the nested expression.
      • ^! (<404684003)
      • ^?MAX (<404684003)
  • ^R 404684003


MRCM on FHIR

All

There is a growing desire to enable access to the SNOMED CT Machine Readable Concept Model via FHIR Terminology Servers.

Use cases:

  • To support the creation and validation of Expression Constraint Language queries
  • Creation and validation of Postcoordinated expressions using Compositional Grammar


How could this work?

Current experiment:

  • An extension of the CodeSystem definition
    • Kai Kewley  drafted a JSON format to represent the required parts of the MRCM to support this use case.
    • Review and discuss
    • FHIR format feedback:
      • Can we split the long string, maybe by domain?
      • Alternative StructureMap rendering?
    • JSON format feedback:
      • Need to represent each attribute-domain
    • Michael Lawley next iteration of the format.


Other options was considered but rejected because of demand of TS implementation complexity:

  • 1. Focus concept(s) -> MRCM Domain → set of domain - attributes
    • ConceptMap? - translate focus concept to set of attributes
      • A map would support returning additional information like grouping
  • 2. Attribute → Search within attribute range
  • New implicit URIs?
  • Consider supporting a reverse lookup - given a focus concept, which attributes can this be used within.

2.3 ECL Enhancement

I want to fetch the set of attribute names used by a concept or set of concepts. 

Use case: 

  • Authoring ECL
  • Authoring template generation

Syntax options:

  • 12334000 . ?
  • 12334000 = R?
  • 12334000 : ?
  • 12334000 [ type ]
  • 12334000.type
  • 12334000 ...

ECL History Supplements with no association

Potential issue: Concepts made inactive with inactivation reason "Non conformance to editorial policy" currently have no historical association.

Use Cases

  • < 414408004 |Hispanic (racial group)| {{ +HISTORY }}
    • No historical associations found against the inactive children

Should ECL be able to retrieve these concepts

Options:

  • Include some association type in RF2 - could be generated before release.
  • Fix in ECL. Terminology server could link concepts without any association to their last active parents.
    • ECL engine could use the snapshot to find the last active parents

Discussed in Joint AG - 

  • Most favoured solution is to add more associations in the RF2 where possible, probably using a scripted process.
    • UK already doing this.
    • TODO: Could Jeremy provide more information on this?

ECL Maintenance Recommendations

All

We plan to add a "Maintenance Recommendations" page to the ECL guide.

Suggested sections:

  • What does best practice ECL look like - what are the benefits
  • ECL comments to record intent
  • Existing tooling and features

Jeremy Rogers has drafted some content on this topic.

Group have reviewed the word document Kai Kewley  to Migrate to ECL Guide Appendix for final review. 

ECL Results - TS consistency

Testing consistency between Snowray (Snow Owl), Ontoserver, Snowstorm. Also using custom made NHS Subset maintenance tool, would like to migrate to a standardised solution.

Questions:

  • Should inactive concepts be routinely suppressed from all results?
    • Answer: Inactive concepts should not be included when using hierarchy constraints, because they are not part of the hierarchy. They should be included when using wildcard. The default substrate includes inactive. Same in FHIR. All agree.
    • Action: Check the spec makes this clear. "Default substrate".
    • Known tooling issues: Snowstorm is not behaving well. e.g. ^ xxx AND YYY (where YYY is inactive). Check both native and FHIR API.

  • Should inactive concepts be returned as active members of a simple refset?
    • Yes - they are active members - all agree
    • Snowstorm issues - seek example

  • Are text definitions and FSNs routinely in scope of any description filters?
    • Text definitions should be left out?
      • Action: change the specification to exclude them by default - done
      • (FHIR should also exclude these when using "filter" parameter)
    • FSN 
      • KK: The FSN can be unique and help find relevant concepts, the PT should be presented
      • JR: The semantic tag can get in the way - 
      • We don't agree on this
      • Action: Let's add more examples around this.

  • Do description filters run by default over inactive descriptions?
    • Inactive are excluded by default in the spec, however there is no way to search active and inactive at once
      • Action: Update ECL Spec to allow searching over active and inactive (apply everywhere, concept, description, members) - done
        • Suggested syntax update ECL 2.3: 
        • .. TS should return error if no inactive descriptions loaded
    • Snow Owl searches active and inactive descriptions by default - needs to be realigned?

  • Is a “pending move (concept)” description active or inactive?
    • JR: Let's skip this for today. It means the concept is about to be moved to another Edition.

  • Are the match and wild description filters case sensitive or not (or, in the case of Ontoserver, do they even work at all?)
    • Group recommends case insensitive matching for "match"
    • Action: Update the ECL guide with this recommendation.
    • Group does not agree about "wild" case sensitivity.
      • Action: Discuss option of adding a modifier to specify case sensitivity - make sure we include Daniel
    • Action: Consider adding ECL regex filter.

  • There are also some interesting edge case differences in the ability of each server to process ECL match description filters if they contain various symbol characters, such as *^% and so on.

    • Action: Discuss updating the ECL spec to add escape character for matching double quotes.
    • MC: Implementation Note - ECL filters commonly include double quotes - when including these in FHIR ValueSet requests make sure they are escaped.
    • Action: Add to ECL Guide and any Snomed FHIR IG.
    • Other characters are perhaps an edge case.
    • More testing needed against know TS.

ML: Some observations: it is interesting to see the number of examples that are querying against terms – the results are not surprising since 1., this is a relatively new part of the spec, and 2., it steps outside the original conception of ECL as a query language that only used defining aspects of concepts as supported by the Concept and Relationship tables.

Postcoordination Guide + Reference Implementation Feedback

Very little feedback received so far.

An education module could be a way to expose more people to the caveats and best practises.

Expression Repository Maintenance

All

- Aim to round off the guidance - make expression repositories feasible -

An expression repository uses a specific edition and version of SNOMED CT - how can expressions be migrated to a new version?

Use cases of expressions:

  • Classifying expressions to discover closest existing precoordinated concept - no requirement for maintenance because expression not recorded in patient record
  • Expression is recorded in patient record - need the ability to find relevant patients. This may not work well over time as SNOMED changes.

Potential issues when upgrading the substrate of an expression repository

  • Inactive concepts within the expression (focus concept, or attribute name or value)
  • Concepts changing hierarchies (semantic tag) - e.g. procedure → observation
    • The has happened for promoted concepts
    • Is this a corner case / rare? Needs analysis. Kai Kewley 
  • Changes to concept model
    • MRCM Changes / Editorial Guide
    • Other modeling changes, perhaps driven by QI consistency efforts
  • Will need to classify again to infer different ancestors and attributes
    • Maybe new equivalents 
  • MC: This has a lot of overlap with ECL maintenance
  • General approaches for identifying issues:
    • FHIR interface could validate the expression again to identify inactive concepts
    • Expand existing value sets to see if there is a radical change in size
    • Content unit tests - needs 
  • Strategies to avoid these issues:
    • Capture user input (close to user form or template slot inputs) so that templates or transformations can be reapplied.
    • Transform input again against the new substrate:
      • Types of template based authoring / transformation
        • Level 1 postcoordinated transformations. 
        • Implementation defined template - input to template captured and stored outside for the terminology server.
        • Precoordinated template pairs - used by International Authoring teams
    • Use the close to user form, level 1 transformations
      • Create expressions using a proximal parent with minimal modeling, lean on Level 1 transformations (see guide)
      • Expression repository should retransform existing close to user form expressions to classifiable expressions
    • Kai Kewley  What techniques from precoordinated authoring could be reused? How do international authors update PPP concepts, what automations?
    • Should expression repository be versioned before being upgraded, or should the repository be recreated against the new substrate?
  • Strategies to fix content issues:
    • Use historical associations to replace inactive concepts used with an expression
      • What historical association types work, which do not?
      • Same As - Switch concept to target
      • Maybe (possibly equivalent to, possibly replaced by)
        • probably can't use both in an expression
        • need a strategy - may not be able to automate, need an author choice?


Use cases for an expression repository, how are these impacted over time as the substrate changes.

  • Postcoordination as a replacement for authoring
    • PPP modeling + attribute
    • Desirable check when upgrading the substrate: detect change in subsumption.

Other Topics

All
  • DK: Discussion of postcoordination transformations for Level 1+ or Level 2
    • use of template language?
    • Transformation choice interactive?
  • JR: History supplements - Deprecate the WAS-A historic association? Causes implementation issues.
    • Bring in other groups? Editorial EAG
The items below are currently on hold
Meds ECL Requirement

Requirement

Requirement to select a substance or a modification of that substance .... use case: during authoring to prompt the user with all the active ingredient options. Author/modeller chooses a drug and wants to refine the selection using modification-of substances. 

  • Alternate use case: Clinician chooses a drug and wants to refine the selection using modification-of substances
    • The product may not exist - it could be made up.
    • B2i have solved this problem using a real-time recursive search

Background

The modification-of attribute is a transitive property in OWL. Transitive properties have a similar characteristic to an is-a relationships but without stating a subtype relationship.

In SNOMED CT an OWL property chains are used to help organise the hierarchy, allowing products with modified substance to be subsumed into concept groupers defined using the base substance. However in this case the base substance should not be inherited into the product with the modified substance. The base (not modified) version of the substance is found as redundant and removed during NNF calculation. See OWL Guide: 2.5. Generating Necessary Normal Form Relationships from the OWL Refsets

Example property chain: "|Has active ingredient|o |Is modification of|" is a sub-property of  "|Has active ingredient|".

Example hierarchy:

Parent

Child

Because

And Property Chain:

SubObjectPropertyOf(
	ObjectPropertyChain(
		:127489000 |Has active ingredient (attribute)|
		:738774007 |Is modification of (attribute)|
	)
	:127489000 |Has active ingredient (attribute)|
)

.. which results in OWL seeing the non modified ingredient as a property of the concept for the purpose of subsumption.

Finally

TransitiveObjectProperty(:738774007 |Is modification of (attribute)|)

This could be used by the ECL engines.


Solution Options:

  • ECL Solution: << 372687004 |Amoxicillin (substance)| 
      OR ( << 105590001 |Substance (substance)| : << 738774007 |Is modification of (attribute)| = << 372687004 |Amoxicillin (substance)| )
      OR ( << 105590001 |Substance (substance)| : << 738774007 |Is modification of (attribute)| = ( << 105590001 |Substance (substance)| : << 738774007 |Is modification of (attribute)| = << 372687004 |Amoxicillin (substance)| ) )
      OR ( << 105590001 |Substance (substance)| : << 738774007 |Is modification of (attribute)| = ( << 105590001 |Substance (substance)| : << 738774007 |Is modification of (attribute)| = ( << 105590001 |Substance (substance)| : << 738774007 |Is modification of (attribute)| = << 372687004 |Amoxicillin (substance)| ) ) )
      )
    • Pros: Already works, if you know the maximum number of transitive hops
    • Cons: It's not pretty. It may work for modification-of but may not work for longer transitive chains like part-of (body structures)


  • NNF Change Solution: Keep the modification-of relationships using a new characteristic type to support querying.
    • Pros: Do not need to increase number of concepts or complexity of ECL engines. 
    • Cons: Big change in the spec? Could be misinterpreted. Seems risky.


  • ECL Engine and Language Change Solution: Build a transitive closure for each transitive property, in the same way that we do for the is-a relationship. We would also need special ECL syntax for this.
    • Pros: No content change. No classification change.
    • Cons: Slightly more work in ECL engines.
    • Pseudo-code: Kai Kewley 
      • In an ECL engine, when building the ECL index:
         
        • Identify which properties are transitive: 
          Set<String> transitiveProperties = owlOntology.getAxioms(AxiomType.TRANSITIVE_OBJECT_PROPERTY).stream()
          .map(axiom -> axiom.getProperty().getNamedProperty().getIRI().getShortForm())
          .collect(Collectors.toSet());

        • Build up a set of transitive property ancestors:
        • Concept:
          Map<String, Set<String>> transitiveTypeParentMap;
          addRelationship(source, destination, type) {
            if (transitiveProperties.contains(type)) {
              transitiveTypeAncestorMap.get(type).add(destination);
            }
          }
        • Include transitive property ancestors in ECL index:
          • conceptId: 351000220100 |Minocycline hydrochloride dihydrate (substance)|
          • ancestors: (usual is-a ancestors)
          • 738774007_ancestors: (52886009 |Minocycline hydrochloride (substance)|, 372653009 |Minocycline (substance)|)
          • 738774007_parents: (52886009)
          • 733930001_ancestors: ()

          • further fields generated depending on the substrate.

      • ECL Language could allow something like: 
        • << [+738774007 |Is modification of|] meaning use modification of transitive hierarchy in addition to is-a hierarchy.
        • or << [+*] meaning use all available transitive hierarchies
        • Examples:
          • < |Product| : << |Has active ingredient| = << [+738774007 |Is modification of|] |Amoxicillin|
          • < |Product| : << |Has active ingredient| = << [+*] |Amoxicillin|

      • JR: At what point do we stop keeping everything stateless and use a reasoner. Should we check our approach / policy?
      • ML: We need a solution that works with all property chains, not just transitive properties... we have that during NNF calculation, we are currently removing it. Perhaps those could be put into the NNF or another file to enable ECL implementations.
      • Kai Kewley and Ale to catchup with Jim and Yong on this.
      • DK: concerned that complexity of ECL would increase and usefulness may decrease? Maybe other architectures are needed to support another way to run snomed queries?
      • JR: Suggests a hybrid ECL engine that can use a reasoner when needed, if the user needs to run a query that requires one and needs the complete answer.
      • ALO: Perhaps SI could pre-compute the ECL index in a standard format for use in any ECL engine. A way to reduce the complexity for implementers and also standardise the solution. This could be a file in the release and also an open source tool.

  • New Type of SNOMED Query Engine including an OWL Reasoner (an idea from Jeremy)
    • Alternative approach to pre-computing the ECL index - see row below.


  • Content Change Solution: Create a grouper substance to include that substance and all single and multi-hop the modification substances.
    • Pros: Would work without other changes. This technique has precedent in the body structure hierarchy.
    • Cons: Approx. three times the number of substance concepts. May be hard for implementers who want a clean list. Because of the high number (around 2K) of substances with a modification relationship these groups would need to be generated.

    • Current content in numbers:



SNOMED OWL Query Engine

Jeremy and all

  • New Type of SNOMED Query Engine including an OWL Reasoner (an idea from Jeremy)
    • Alternative approach to pre-computing the ECL index. The engine would detect when OWL reasoning is required and use OWL to execute specific types of query.. for example queries involving property chains where the NNF is not sufficient.
    • How is this different from postcoordination.
    • What is the gap between ECL and Postcoordination:
      • ECL things not available in PCE engines:
        • ECL filters
        • Refset membership and other fields
        • Cardinality constraints
    • Use Cases:
      • Clinical use case: Refining the selection of a drug (using property chains)
      • Back office / QA use case: Find all concepts with only one parent.
        • B2i support this by extending the spec and allowing cardinality on the is-a relationship.
      • Maybe in the future other attributes will be transitive: caused-by, due-to.



Attachments

...