W3C-HCLS-COI/HL7-ITS-RDF joint call

04 Nov 2014

See also: IRC log


Lloyd_McKenzie, Kerstin_Forsberg, Rafael, DaleNelson_, PKnapp, Rob_Hausam, mscottm, Guoqian, Sal, Tony, Josh, Claude_Nanjo, ratnesh, mehmet, egombocz, Vipul, RRSAgent, Souri, hcools, TallTed, cloudcell, ericP_Souri, Josh, Bryn, RafaelRichards, Tony, LaurieWoodrome
Claude_Nanjo, ericP


logistics intro

<Claude_Nanjo> Good morning everyone

<ratnesh> ??P24 is Ratnesh

<Marc_AgfaHealthCare> Hello There

<Marc_AgfaHealthCare> this is Marc from Belgium

<Marc_AgfaHealthCare> having problem to connect to TCon... any change about the conference code?

<Vipul> +1.617.498.aaaa is Vipul

<Marc_AgfaHealthCare> conference code?

<Marc_AgfaHealthCare> succeded to join the call

<ericP> DaleNelson: ITS focuses on physical rendering. We don't promote models of our own.

<ericP> PaulKnapp: HL7 quarum requirements require an ITS co-chair. we'll step in to chair for motions and give back to the regular chair

<ericP> Claude_Nanjo: I've been working with David Booth.

<ericP> ... This is an opp for the HL7 and W3C communities to exchange knowledge and ideas.

<ericP> ... David will be hosting many of the calls.

<Sal> What is the number to call in?

Clinical Observation & Interoperability Task Force - coding medical records Goal - W3C and HL7 can complement each other's work W3C Work - Technical aspect of building a clinical infrastructure using semantic technologies - Clinical care/trials can really benefit from semantic technologies HL7 (Paul Knapp & Dale Nelson) - ITS is focused on physical rendering of content models and in the transport of information - No content of their own - Techno[CUT]

<ericP> Sal, +1.617.761.6200 4257#

... for exchanging the content of others. RDF fits fairly well within this scope. - Minimum criteria - One co-chair and couple of attendees. - Decision must follow HL7 process - As long as HL7 co-chair is present, decisions can be made within this group. If not, they need to be brought back to ITS. - Dale Nelson - introduction. - Quorum - at least one co-chair needed for formal motions and votes - Claude - introduction

<Sal> Thanks!

<Claude_Nanjo> Paul please feel free to correct the information if not accurate.

Minutes to provide of record of decisions made and inform the wider group.

<Claude_Nanjo> Format:

<Claude_Nanjo> Person's name: ...

<ericP> scribenick: Claude_Nanjo

No web screen share but we can arrange for one if needed.

FHIR ontology

<ericP> scribenick: ericP

Claude_Nanjo: we started some work on a FHIR ontology

<Marc_AgfaHealthCare> Is there any material to share on the work done about FHIR ontology?

Claude_Nanjo: we alligned with an ontology that Josh Mandel had done to present FHIR as RDF
... we paused 4-5 months ago
... the goal was to find a SemWeb representation
... discussed with Lloyd and JoshM
... some open issues around the way FHIR models concepts which could be represented in another way
... and how to link to other ontologies and use in other contexts

<ratnesh> http://gforge.hl7.org/gf/project/hl7owl/

ratnesh: there is work at HL7 on O-RIM.
... we could map between FHIR and O-RIM, especially the datatypes.

Claude_Nanjo: there were discussions in CDS around a shallow hierarchy for FHIR.
... I didn't add a hierarchy to FHIR's root-level resources.
... We didn't want to make the structures unintuitive for the SemWeb community.

lloyd: we should be leary about using O-RIM to drive the FHIR hierarchy (that we choose to expose).
... what the RIM ontology might impose may or may not be relevent from a business perspective.
... There is a desire to have a formal representation of FHIR instances that can be transformed to/from the other two syntaxes.
... That would fall under HL7 ITS and would be appropriate for this group. It's been on the FHIR wish list for a while.

ericP: JoshM and I worked on stuff to map FHIR XML <-> RDF
... the reference implementaitons in C#, Java, etc. can be updated to R/W RDF

GuoqianJiang: at Mayo, we're using ISO11179 to capture structures from FHIR because we want to capture data elements from different models.
... we want to capture e.g. QDM, CEMs from IMH
... The RDF version of FHIR would help us


Rafael: Mark Musen will be giving a talk soon.
... VA has put our vocab into SKOS.
... good exposure of Vista in RDF via Connor's work.
... so VA is trying to leverage this large repo

Claude_Nanjo: next week Mark will present on this topic.

Bridging VA, Intermountain and other models

Claude_Nanjo: this isn't a current project.
... there are a number of models out there (FHIR, VMR, IMH's work on CIMI, VA models...)
... we discussed a way to bridge these models.
... e.g. expose DOD patient model info as FHIR
... transform RDF patient graphs into FHIR
... we are normalizing and doing analytics so we need the original structure.
... can we link these models?

Use cases

<scribe> scribenick: Claude_Nanjo

GuoqianJiang: how do we capture use cases?

Guokian Jiang: It would be good to identify use cases that connect this work. It would be good to have a wiki to do this.

<Marc_AgfaHealthCare> sorry I will need to leave the Tcon

<Marc_AgfaHealthCare> will catch up with the minute

Eric: We could use the W3C wiki. May need some governance to identify similar work.

<ericP> ACTION: ericP to send pointers to existing wiki pages [recorded in http://www.w3.org/2014/11/04-hcls-minutes.html#action01]

Eric will provide pointers to existing Wiki

PhUSE-FDA project (formerly CDISC2RDF)

C-CDA RDF representations

<ericP> Turtle representaiton of a C-CDA

Eric: Introduced a framework for the conversion of CCDA to turtle.

Rafael: Could this also be outputted as JSON-LD?

Eric: May be a fairly nested structure if represented as JSON-LD

PhUSE-FDA project (formerly CDISC2RDF)

Rafael: May be friendlier to front end developers

High-level concept mapping to RDF (AR typeCodes, etc.)

Tony: convergence is a high priority or we end up in an entropy of opinions

Tony: Convergence is high on our list. Assuming we get to some level of convergence, and say it is FHIR, we may need to do some high level mappings of concepts to RDF. How do terminology codes map to RDF - looked at IHTSDO for how concepts look like in this ontology.

<ericP> slice of SNOMED-CT

How does a coded term look like in RDF, what is its URI, its URI structure? We should take on high level mappings of high level models to RDF. Can we converge on styles of representation.

<ericP> slice of SNOMED-CT with descriptive URLs

Tony: How does one relate a concept to an instance

GuokianJiang: IHTSDO and WHO are trying to harmonize

IHTSDO and WHO are discussing convergence?

... They have had an agreement to develop a common ontology for ICD-11 and SNOMED. T

<Guoqian> The idea is that each ICD-11 category will have a formal definition from SNOMED CT concepts

Lloyd and Tony: Concept may be represented with coding from multiple terminologies and needing to identify code, code system, version (URI) to associate the instance to the concept. A constraint: What is done in RDF will need to roundtrip from RDF to XML/JSON. What is the scheme to represent these identifiers.

Summary of Action Items

[NEW] ACTION: ericP to send pointers to existing wiki pages [recorded in http://www.w3.org/2014/11/04-hcls-minutes.html#action01]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.138 (CVS log)
$Date: 2014/11/04 17:10:28 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.138  of Date: 2013-04-25 13:59:11  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Found ScribeNick: Claude_Nanjo
Found ScribeNick: ericP
Found ScribeNick: Claude_Nanjo
Inferring Scribes: Claude_Nanjo, ericP
Scribes: Claude_Nanjo, ericP
ScribeNicks: Claude_Nanjo, ericP
Present: Lloyd_McKenzie Kerstin_Forsberg Rafael DaleNelson_ PKnapp Rob_Hausam mscottm Guoqian Sal Tony Josh Claude_Nanjo ratnesh mehmet egombocz Vipul RRSAgent Souri hcools TallTed cloudcell ericP_Souri Josh Bryn RafaelRichards Tony LaurieWoodrome

WARNING: No meeting title found!
You should specify the meeting title like this:
<dbooth> Meeting: Weekly Baking Club Meeting

Got date from IRC log name: 04 Nov 2014
Guessing minutes URL: http://www.w3.org/2014/11/04-hcls-minutes.html
People with action items: ericp

[End of scribe.perl diagnostic output]