CIMI WGM Agenda May 2016

Return to Clinical_Information_Modeling_Initiative_Work_Group main page

Agenda

Day	Time		Room	Event	Host	Joining	Chair	Scribe
Sunday May 8	AM	Q1		Administration (week planning, wg health, project status updates)
		Q2		Governance, what's in the modeling pipeline, how are models approved, what goes in the repository			Stan
	PM	Lunch		How when and why are model transformations done (isosemantic, translation to design models)			Claude, Richard
		Q3		Design - composition			Claude
		Q4		Design - Alignment with FHIM			Galen
Monday May 9	AM	Q1		Tooling: ISAAC, IHTSDO, 'archetype design tool'			Craig, Keith, Harold, Linda
		Q2		Semantics & alignment with models of meaning, e.g., SNOMED CT; binding; handling federated terminologies; selection of concept identifiers (e.g., LOINC vs Observable)			Jay, Claude, Keith
		lunch		Semantics & alignment with IHTSDO			Linda
	PM	Q3		Quality Metrics for DCMs, incl. ISO			Sun-Ju
		Q4		Tooling 2: SCT CIMI identifer tool, knowledge management, experimental integration with CQL			Harold, Richard
Day	Time		Room	Event	Host	Joining	Chair	Scribe
Tuesday May 10	AM	Q1	Salon 6 & 7	@ Patient Care, skin breakdown	Patient Care	CIMI	Jay, Susan, Harold
		Q2	Salon Drummond Centre	Argonauts			Stan
	PM	lunch		No meeting. Previously CIMI meeting with FHIR Core Team
		Q3	Salle De Bal East	At Vocabulary	Vocabulary	CIMI
		Q4	Salon 6 & 7	NOT MEETING Previously scheduled with PC, EC, FHIR, Voc
Day	Time		Room	Event	Host	Joining	Chair	Scribe
Wednesday May 11	AM	Q1	Salon Hemon	@ CQI	CQI	CIMI	Claude, Bryn, Ken
		Q2	Salon Hemon	FHIR semantics; representing archetypes as structure definitions			Grahame
	PM	Q3	Salon Hemon	RM (status); AML to ADL transforms (i.e., FHIM); possible RDF support			Michael
		Q4	Salle De Bal East		CIMI	CQI and CDS	Claude, Bryn, Ken
Day	Time		Room	Event	Host	Joining	Chair	Scribe
Thursday May 12	AM	Q1	Suite 701	@ CIC DAM Data Elements	CIC	CIMI
		Q2		NOT MEETING
	PM	Q3		NOT MEETING
		Q4		NOT MEETING
Day	Time		Room	Event	Host	Joining	Chair	Scribe
Friday May 13	AM	Q1		NOT MEETING
		Q2		NOT MEETING
	PM	Q3		NOT MEETING
		Q4		NOT MEETING

Minutes

Sunday Q3

HL7 CIMI Meeting Agenda/Minutes
Location: Room xyz	Date: 2016-01-10 Time: 11:00 AM Central
Facilitator:	Note taker(s):
Attendee	Name, Affiliation
.	.
Quorum Requirements Met (co-chair plus 3 counting staff):

Minutes/Conclusions Reached:

Sunday Q4

Attendees Michale Van der Zel, Jay Lyle, Chris Millet, Floyd Eisenberg, Susan Matney, Craig Parker, Joey Coyle, Stan Huff, Linda Bird, Harold Solbrig, Claude Nanjo, Richard Esmond, 3-6 others

• Jay presented the FHIM at a high level, including intended uses and a brief view of the model.
• The expectation is that FHIM could provide a framework for inserting CIMI components.
• The question of composition was addressed. A CIMI indivisible model must provide the context necessary for its interpretation, so logically, a panel would contain repetitive context information, but an implementation could use pointers to avoid excessive repetition.
• We need to address the compositional dimensions Claude brings up; they seem to be orthogonal to the compositional concept of FHIM/CIMI
• Question: is this specialization or constraint? Does it matter?
• We need an approach for templating (localizing) archetypes
• Stan to provide a list of questions already identified for task force

Monday Q2

• Claude and Harold presented slides on the CQI approach to model composition and on AML representation of terminology bindings.
• Linda presented ongoing work on IHTSDO syntax development
• Jay presented slides on a project to use xslt to transform archetypes into owl xml. Questions included the following:
  • Maturity of CIMI models: actually, current drafts are based on extensive harmonization, and model bindings to SCT may be incomplete but are vetted. What we are actually missing is a clear governance path so you can tell what's official and what's in process.
  • Maturity of Observable model: a new version is available, or is to be available.
  • Value set binding syntax: it's a URI. How this is to be resolved is the responsibility of a service, not CIMI.
  • Refset bindings will require a service in order to create closure axioms for owl.
  • XML version of CIMI archetypes is not normative. This one we should discuss.