Re: Propose an HL7 work group on RDF for Semantic Interoperability?

Strictly speaking, this is not an “official” mapping, and it is subject to change without warning.  However, it is not unreasonable to use it as indicative of the DL representation of SNOMED CT and it was the model used when the IHTSDO moved their tooling from a bespoke algorithm to Snorocket, which implements EL+.

Now that the SNOMED CT URI Specification has been adopted by IHTSDO, one of my goals as Technical Committee member is to formalise an OWL representation for SNOMED CT.

As it stands today, SNOMED CT has role hierarchies and one role chain – a right identity, but last I looked this was actually unused.

michael

—
Michael J Lawley, PhD
Principal Research Scientist | Health Information Environment Project
The Australian e-Health Research Centre
CSIRO

Phone: +61 7 3253 3609 | Fax: +61 7 3253 3690 | Mobile: 0427 456 260
Michael.Lawley@csiro.au | www.csiro.au | www.aehrc.com/hie
Address: Level 5 - UQ Health Sciences Building 901/16, Royal Brisbane and Women's Hospital, Herston, QLD 4029 Australia

PLEASE NOTE
The information contained in this email may be confidential or privileged. Any unauthorised use or disclosure is prohibited. If you have received this email in error, please delete it immediately and notify the sender by return email. Thank you. To the extent permitted by law, CSIRO does not represent, warrant and/or guarantee that the integrity of this communication has been maintained or that the communication is free of errors, virus, interception or interference.

From: "Peter.Hendler@kp.org<mailto:Peter.Hendler@kp.org>" <Peter.Hendler@kp.org<mailto:Peter.Hendler@kp.org>>
Date: Friday, 16 May 2014 9:48 am
To: "mccusj@rpi.edu<mailto:mccusj@rpi.edu>" <mccusj@rpi.edu<mailto:mccusj@rpi.edu>>
Cc: "amallia@edmondsci.com<mailto:amallia@edmondsci.com>" <amallia@edmondsci.com<mailto:amallia@edmondsci.com>>, "cnanjo@cognitivemedicine.com<mailto:cnanjo@cognitivemedicine.com>" <cnanjo@cognitivemedicine.com<mailto:cnanjo@cognitivemedicine.com>>, "david@dbooth.org<mailto:david@dbooth.org>" <david@dbooth.org<mailto:david@dbooth.org>>, "eric@w3.org<mailto:eric@w3.org>" <eric@w3.org<mailto:eric@w3.org>>, "michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com>" <michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com>>, "public-semweb-lifesci@w3.org<mailto:public-semweb-lifesci@w3.org>" <public-semweb-lifesci@w3.org<mailto:public-semweb-lifesci@w3.org>>, "Rafael.Richards@va.gov<mailto:Rafael.Richards@va.gov>" <Rafael.Richards@va.gov<mailto:Rafael.Richards@va.gov>>
Subject: Re: Propose an HL7 work group on RDF for Semantic Interoperability?
Resent-From: <public-semweb-lifesci@w3.org<mailto:public-semweb-lifesci@w3.org>>
Resent-Date: Friday, 16 May 2014 9:49 am

Yes there is a formal mapping. If you download SNOMED from the NLM there will be a PERL script written by Kent Spackman somewhere in there. It will use the official release files and generate an OWL file. That is the official mapping.  The Plus in EL+ I used to know. I think it means hierarchical roles and role groups.  Or maybe it was role chaining. It was one of those things with roles.


NOTICE TO RECIPIENT:  If you are not the intended recipient of this e-mail, you are prohibited from sharing, copying, or otherwise using or disclosing its contents.  If you have received this e-mail in error, please notify the sender immediately by reply e-mail and permanently delete this e-mail and any attachments without reading, forwarding or saving them.  Thank you.




From:        Jim McCusker <mccusj@rpi.edu<mailto:mccusj@rpi.edu>>
To:        Peter Hendler/CA/KAIPERM@KAIPERM
Cc:        amallia@edmondsci.com<mailto:amallia@edmondsci.com>, cnanjo@cognitivemedicine.com<mailto:cnanjo@cognitivemedicine.com>, David Booth <david@dbooth.org<mailto:david@dbooth.org>>, "Eric Prud'hommeaux" <eric@w3.org<mailto:eric@w3.org>>, Michel Dumontier <michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com>>, w3c semweb HCLS <public-semweb-lifesci@w3.org<mailto:public-semweb-lifesci@w3.org>>, Rafael.Richards@va.gov<mailto:Rafael.Richards@va.gov>
Date:        05/15/2014 03:04 PM
Subject:        Re: Propose an HL7 work group on RDF for Semantic Interoperability?
________________________________



Is there a formal mapping (and maybe impelementation?) between SNOMED and OWL EL+? What's been added to the +?

Thanks,
Jim


On Thu, May 15, 2014 at 5:34 PM, <Peter.Hendler@kp.org<mailto:Peter.Hendler@kp.org>> wrote:
On the last point. There is a post coordination syntax for SNOMED (which if you understand OWL is really very similar with different syntax).  I'd think for the last use case you might as well use the SNOMED post coordination syntax. It's actually easier than most representations of OWL, and anyone who knows enough SNOMED to post coordinate will certainly know this syntax.

For point one, I guess if you wanted to parse your FIHR into a triple store that might be nice, but FIHR itself isn't made with ontology in mind.  Entities in Roles that Participate in Acts work wonderfully in OWL, But Person FIHR resources aren't about those kind of relationships.

Point two. SNOMED is EL+ which is a subset of OWL DL so this can be done, but again, the added relationships in FIHR aren't very helpful, and besides, EL+ does not scale to millions of patients worth of data for a reasoner (it does for SPARQL).  I suppose there is a use case for that..  But then it doesn't have to be OWL, it can just be RDF triples.


OWL-2RL is massively scalable, but is a different flavor or logic than SNOMED.  I like the idea of keeping SNOMED in EL+ and if you do add clinical information, then make it Entities in Roles Participating in Acts and make it only in OWL-2RL.


NOTICE TO RECIPIENT:  If you are not the intended recipient of this e-mail, you are prohibited from sharing, copying, or otherwise using or disclosing its contents.  If you have received this e-mail in error, please notify the sender immediately by reply e-mail and permanently delete this e-mail and any attachments without reading, forwarding or saving them.  Thank you.




From:        Anthony Mallia <amallia@edmondsci.com<mailto:amallia@edmondsci.com>>
To:        Peter Hendler/CA/KAIPERM@KAIPERM, "michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com>" <michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com>>
Cc:        "david@dbooth.org<mailto:david@dbooth.org>" <david@dbooth.org<mailto:david@dbooth.org>>, "eric@w3.org<mailto:eric@w3.org>" <eric@w3.org<mailto:eric@w3.org>>, "cnanjo@cognitivemedicine.com<mailto:cnanjo@cognitivemedicine.com>" <cnanjo@cognitivemedicine.com<mailto:cnanjo@cognitivemedicine.com>>, "Rafael.Richards@va.gov<mailto:Rafael.Richards@va.gov>" <Rafael.Richards@va.gov<mailto:Rafael.Richards@va.gov>>, "public-semweb-lifesci@w3.org<mailto:public-semweb-lifesci@w3.org>" <public-semweb-lifesci@w3.org<mailto:public-semweb-lifesci@w3.org>>
Date:        05/15/2014 01:53 PM
Subject:        RE: Propose an HL7 work group on RDF for Semantic Interoperability?
________________________________



Peter,
Good to see you at the HL7 meeting.
There seem to be three approaches being worked in RDF

1.       The first is expressing the FHIR payload in RDF. The ontology is therefore the FHIR ontology.. The discussion at HL7 was about making another representation of the payload in RDF.

2.       The second (which I have been working on) has a small sample in Protégé is to look at the ontologies of FHIR and Terminologies – particularly SNOMED CT and express them together in the same language – OWL so that the conflicts can be resolved. The binding from objects in FHIR to terms is fundamentally different from the FHIR payload and therefore more disruptive. In this approach, a higher ontology is critical – the RIM was a start but has quite a lot missing to form a good higher ontology and has nothing to do with healthcare (The RIM did not either). It looks as if you are going in this direction as well.

3.       The third which comes from the terminology side is to express post-coordinated terms as a code OWL expression in a FHIR message. This is a small scope but seems to have some importance to resolving the terminology issues.

These three approaches do not overlap or conflict as far as I can see.

Maybe there are two projects one which is aligned to represent FHIR and one which looks at the broader issue of Healthcare Ontology as you have described. I don’t think the third deserves a project.

Tony Mallia

From: Peter.Hendler@kp.org<mailto:Peter.Hendler@kp.org> [mailto:Peter.Hendler@kp.org]
Sent: Tuesday, May 13, 2014 5:33 PM
To: michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com>
Cc: david@dbooth.org<mailto:david@dbooth.org>; Anthony Mallia; eric@w3.org<mailto:eric@w3.org>; cnanjo@cognitivemedicine.com<mailto:cnanjo@cognitivemedicine.com>; Rafael.Richards@va.gov<mailto:Rafael.Richards@va.gov>; public-semweb-lifesci@w3.org<mailto:public-semweb-lifesci@w3.org>
Subject: Re: Propose an HL7 work group on RDF for Semantic Interoperability?

We are in fact doing just that at Kaiser on a small experimental scale. We are working with Oxford and RDFox and creating an HL7 "RIM Like model".  It is not the full RIM but is Entities in Roles Participating in Acts.

I think it would be a mistake to try to put the whole RIM into RDF.  Also I'm not a fan of a de novo RDF model that is not based on Entities in Roles Participating in Acts, It would then just be an arbitrary other model.  You could put FIHR to RDF but I see little reason for it.  Since RDF is wide open (almost like saying lets do something in XML) you really have to think about what the model is based on before you start.  You've only got V2 (no good for RDF) RIM (much too big for RDF) and FIHR (not much advantage to RDF for this).

With our "RIM like" model based on Entities in Roles that Participate in Acts, we expect to be able to run useful safety and outcome reports.  Also we use SNOMED so the ontology of SNOMED will be tightly coupled to our model, but our model will not be an extension of the SNOMED model.

I'd be disappointed to see HL7 create a brand new model in RDF that is not based on Entities in Roles Participating in Acts.





NOTICE TO RECIPIENT:  If you are not the intended recipient of this e-mail, you are prohibited from sharing, copying, or otherwise using or disclosing its contents.  If you have received this e-mail in error, please notify the sender immediately by reply e-mail and permanently delete this e-mail and any attachments without reading, forwarding or saving them.  Thank you.




From:        Michel Dumontier <michel.dumontier@gmail.com<mailto:michel.dumontier@gmail.com>>
To:        David Booth <david@dbooth.org<mailto:david@dbooth.org>>
Cc:        Anthony Mallia <amallia@edmondsci.com<mailto:amallia@edmondsci.com>>, "Prud'hommeaux, Eric" <eric@w3.org<mailto:eric@w3.org>>, Claude Nanjo <cnanjo@cognitivemedicine.com<mailto:cnanjo@cognitivemedicine.com>>, "Richards, Rafael M. (Rafael.Richards@va.gov<mailto:Rafael.Richards@va.gov>)" <Rafael.Richards@va.gov<mailto:Rafael.Richards@va.gov>>, w3c semweb HCLS <public-semweb-lifesci@w3.org<mailto:public-semweb-lifesci@w3.org>>
Date:        05/13/2014 02:01 PM
Subject:        Re: Propose an HL7 work group on RDF for Semantic Interoperability?


________________________________





David,
 I'm excited about the prospect, but I think the work group may be a bit too broad without further refinement. I'd like to see a charter articulated with a more specific focus, and identifying milestones that deliver concrete outcomes (specifications, implementations, reports) around targeted areas of urgent need.
More importantly, I'd like to know what the uptake will be (e.g. who will implement this), and how plans on using it.

Cheers,

m.


Michel Dumontier
Associate Professor of Medicine (Biomedical Informatics), Stanford University
Chair, W3C Semantic Web for Health Care and the Life Sciences Interest Group
http://dumontierlab.com<http://dumontierlab.com/>


On Tue, May 13, 2014 at 7:54 AM, David Booth <david@dbooth.org<mailto:david@dbooth.org>> wrote:
I was at the HL7 meetings last week, and the idea arose of proposing an HL7 work group on RDF for Semantic Interoperability.  I took the initiative to draft a possible charter.  I meant to send it out earlier in case folks would like to discuss the idea on our 11:00am Eastern HCLS call today.  Attached is what I've drafted.  I'll join the call and see if anyone wants to discuss it.  Sorry for the late notice.

Zakim (W3C teleconference bridge).
Dial-In #: +1.617.761.6200<tel:%2B1.617.761.6200> (Cambridge, MA)
VoIP address: sip:zakim@voip.w3.org<mailto:sip%3Azakim@voip.w3.org>
Participant Access Code: 4257 ("HCLS")
IRC: irc.w3.org<http://irc.w3.org/> port 6665 channel #HCLS

Thanks,
David




--
Jim McCusker

Data Scientist
5AM Solutions
jmccusker@5amsolutions.com<mailto:jmccusker@5amsolutions.com>
http://5amsolutions.com<http://5amsolutions.com/>

PhD Student
Tetherless World Constellation
Rensselaer Polytechnic Institute
mccusj@cs.rpi.edu<mailto:mccusj@cs.rpi.edu>
http://tw.rpi.edu<http://tw.rpi.edu/>