15:01:56 <RRSAgent> RRSAgent has joined #hcls
15:01:56 <RRSAgent> logging to http://www.w3.org/2015/09/08-hcls-irc
15:01:58 <trackbot> RRSAgent, make logs world
15:01:58 <Zakim> Zakim has joined #hcls
15:02:00 <trackbot> Zakim, this will be HCLS
15:02:00 <Zakim> I do not see a conference matching that name scheduled within the next hour, trackbot
15:02:01 <trackbot> Meeting: Semantic Web Health Care and Life Sciences Interest Group Teleconference
15:02:01 <trackbot> Date: 08 September 2015
15:05:03 <dbooth> Topic: FHIR ValueSets
15:05:11 <dbooth> http://wiki.hl7.org/index.php?title=File:Vocabulary_and_Terminology_in_RDFv2.pdf
15:05:15 <Tony> Tony has joined #HCLS
15:06:21 <Lloyd> Lloyd has joined #HCLS
15:09:28 <dbooth> http://wiki.hl7.org/index.php?title=File:Comparison_of_ValueSet_approaches.pdf
15:10:22 <dbooth> lloyd: approach 1 is how it is in FHIR.
15:11:07 <dbooth> lloyd: approach 2 is a subset of 1.  some valuesets will make use of the concept hierarchy.
15:11:26 <dbooth> ... That's a common approach, but not the only one.
15:11:56 <dbooth> tony: But there are no concepts in #1.
15:12:23 <dbooth> lloyd: But concepts can be used with #1.   So #2 is an option with #1, but you cannot get #2 out of #1.
15:13:08 <dbooth> rob: We're not complete separating valueset definitions and valueset expansions.  You define the VS and you can do it how you want, based on codes, or descendents of concept X, etc.
15:13:32 <dbooth> ... Then you process and expand it to get the actual list of codes, and that's what you work with.
15:14:03 <dbooth> lloyd: Expansion is a pointing time list that gets the code by evaluating the expressions that define what codes are included.
15:14:33 <dbooth> ... We're not talking about the clinical fact at all.
15:14:58 <dbooth> ... If I evaluate an expression now vs two years ago, i'll get two different sets of codes.
15:15:43 <dbooth> tony: We bound to the medical fact when we bound the type to the observation instance.
15:16:57 <dbooth> lloyd: There are two things going on.  In the FHIR world, whenever you pass a resource around, you're passing around an instance of a resource.
15:17:18 <dbooth> ... When you turn that into ont space, (only with certain resources), they become classes.
15:17:51 <dbooth> ... A VS instance (e.g in XML  )  exists independent of a medical fact.
15:18:08 <dbooth> ... The expansion of the codes can be generated independent of medical facts also.
15:19:19 <dbooth> ... The expansion operation takes a VS instance and a date and produces a new VS instance with the expansion portion filled in.
15:19:41 <dbooth> ... And the expansion is a collection of Codings.
15:21:09 <dbooth> tony: on line 24, in the allergy-intolerance-status class, the individual is saying 'here is a restriction on something that can be allowed or not'
15:25:30 <dbooth> lloyd: we define our own URLs for all SNOMED concepts.
15:26:05 <dbooth> q+
15:34:03 <dbooth> ack dbooth
15:34:40 <dbooth> ACTION: Tony to add FHIR XML examples to ValueSet document
15:34:40 <trackbot> Error finding 'Tony'. You can review and register nicknames at <http://www.w3.org/2014/HCLS/track/users>.
15:36:48 <dbooth> ACTION: Tony to change lines 24-37 to show codes *allowed* in allergy-intolerance-status
15:36:48 <trackbot> Error finding 'Tony'. You can review and register nicknames at <http://www.w3.org/2014/HCLS/track/users>.
15:40:44 <dbooth> dbooth: lines 46-56 defines fhirvs:substance-code to be a single system/code pair.
15:41:11 <dbooth> tony: Yes.  it needs the transitive concept-is-a added, to get all codes that are sub-codes.
15:48:04 <dbooth> dbooth: lines 120-121 look like they need to be in the snomed bridging ontology -- not in the generic FHIR ont.
15:48:39 <dbooth> lloyd: bridging ont also needs to fill in things that are missing.
15:49:05 <dbooth> ... Eg, doing a bridge to LOINC, and using specific names for properties.
15:50:13 <dbooth> ... And official LOINC ont does not expose it, or exposes it using a different name.  The bridging ont would need to expose all of the pieces that are needed for using the base ont with FHIR.
15:50:31 <dbooth> ... What is base ont?  Loinc publishes ont for its codes.
15:51:29 <dbooth> ... The bridge ont will take the aliases that we have defined, and link them to LOINC URLs.  Could also define relationships, properties, etc. that were not exposed or with the expected URIs.
15:52:21 <dbooth> rob: agree.  it defines things beyond what's in the LOINC ont.
15:54:40 <dbooth> dbooth: Need to show: 1. FHIR XML; 2. FHIR RDF that is produced with no knowledge of SNOMED ont; 3. relevant exerpt of SNOMED ont; 4. FHIR-SNOMED bridge ont.
16:02:23 <dbooth> lloyd: I think we really have one approach with two options
16:04:12 <dbooth> ADJOURNED
16:04:46 <dbooth> Present: Rob Hausam, Tony Mallia, David Booth, Lloyd McKenzie
16:04:55 <dbooth> Chair: David Booth
16:05:07 <dbooth> rrsagent, draft minutes
16:05:07 <RRSAgent> I have made the request to generate http://www.w3.org/2015/09/08-hcls-minutes.html dbooth
16:08:40 <MacTed> MacTed has joined #hcls
18:26:28 <Zakim> Zakim has left #hcls