RE: Our white paper on Semantic Node Labeling in Clinical models

Very nice paper, Peter, thanks for sharing.

The more I’ve been thinking about it, the more I come to the conclusion there isn’t a boundary between information model and terminology. There is a continuum.
The only solution is , as proposed, to ensure that the information model is based on terminology.
Then only the information captured is extensional, and modeling becomes maximally intensional.

The “systolic blood pressure” is a nice example.
Systems need to be able to align the observation “systolic blood pressure elevated” with the observable entity “systolic blood pressure” and the procedure “blood pressure taking” or, if it’s a patient reported measure, “blood pressure reporting”.

I truly hope we can raise the bar on this topic.

Cheers, Ronald


From: Peter.Hendler@kp.org [mailto:Peter.Hendler@kp.org]
Sent: dinsdag 18 september 2012 16:13
To: ratnesh.sahay@deri.org
Cc: eric@w3.org; helena.deus@deri.org; kerstin.l.forsberg@gmail.com; LINMD.SIMON@mcrf.mfldclin.edu; meadch@mail.nih.gov; mscottmarshall@gmail.com; public-semweb-lifesci@w3.org
Subject: RE: Our white paper on Semantic Node Labeling in Clinical models

I think I know what the difference is between our views.

I agree RDF might have been a better way to go then regular OO UML models.  But I'm addressing the state of the clinical modeling world for the past 10 years or more.
The three main world wide clinical modeling systems in use are HL7 RIM, openEHR, and ISO 13606.
The vast majority (maybe even all) clinical models from standards organizations are OO, and most of them sometimes use SNOMED and often mess up the boundary between the
intensional and extensional parts of the models.

Our paper is to correct the current state of clinical modeling as it is and has been for many years.

It might be that these issues would be addressed differently if we were talking about a mixture of RDF and SNOMED.  But that's not how it is out in the world.

I wonder if there were mappings between the OO extensional parts of HL7 CDA, openEHR, ISO 13606 and an RDF representation what new things could be done.
Probably a lot of useful new queries.

We are pointing out the current state of the clinical modeling world of the three International standards currently used, and the mixing up of SNOMED with them.

Not the theoretical situation of mixing RDF and SNOMED which might be very useful and interesting.




[cid:image002.jpg@01CD966F.F04DA8D0]



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.




"Sahay, Ratnesh" <ratnesh.sahay@deri.org<mailto:ratnesh.sahay@deri.org>>

09/18/2012 04:15 AM

To

Peter Hendler/CA/KAIPERM@KAIPERM

cc

<eric@w3.org<mailto:eric@w3.org>>, "Deus, Helena" <helena.deus@deri.org<mailto:helena.deus@deri.org>>, <kerstin.l.forsberg@gmail.com<mailto:kerstin.l.forsberg@gmail.com>>, <LINMD.SIMON@mcrf.mfldclin.edu<mailto:LINMD.SIMON@mcrf.mfldclin.edu>>, <meadch@mail.nih.gov<mailto:meadch@mail.nih.gov>>, <mscottmarshall@gmail.com<mailto:mscottmarshall@gmail.com>>, <public-semweb-lifesci@w3.org<mailto:public-semweb-lifesci@w3.org>>

Subject

RE: Our white paper on Semantic Node Labeling in Clinical models







Hi Peter,

I agree that extensional (closed world clinical models) and intentional (SNOMED) models need a separate treatment. Developers (human)  or reasoners (machine) need a mechanism to understand their boundary and act appropriately. For example, below annotations in RED would help developers to interpret their scope of use.

--OWL Manchester Syntax---

(1 ) DataProperty: effectiveTime
                Domain: ActLabObservation (*coming from HL7 RIM*)
                Range: xsd:dateTime
                Annotations: rdfs:label "local or extentional property"


(2) ObjectProperty: snomed:DirectSubstance
                Annotations: rdfs:label "Substance on which the lab procedure method directly acts"
                Annotations: rdfs:label "global or intentional property"
SubPropertyChain:
                snomed:DirectSubstance o snomed:HasActiveIngredient

However,  annotations (or lables) are reasoner-insensitive, so the separation and use of both type of models (extensional, intentional)  is entirely on a developer’s understanding.  IMO, to avoid “undesirable inferences” by “mixing-up” these two types of models (extensional, intentional), caution should be taken at a stage where alignments (subclass, equivalentClass, etc.) are created between them.  Also, there  should be a guideline (for specific settings) about (1) when interfaces are required (i.e., reasoners); and (2) when only querying (SPARQL) is sufficient. This will help further on your proposal for SNL. Since long there has been a discussion on this [1,2] in the Semantic Web Community, still we need to see something concrete.

[1] B. Motik, I. Horrocks, R. Rosati, and U. Sattler, “Can OWL and Logic Programming Live Together Happily Ever After?” in International Semantic Web Conference, 2006, pp. 501–514. [Online]. Available: http://www.cs.ox.ac.uk/boris.motik/pubs/mhs06happily.pdf


[2] B. Motik, I. Horrocks, and U. Sattler, “Bridging the gap between OWL and relational databases,” J. Web
Sem., vol. 7, no. 2, pp. 74–89, 2009. [Online]. Available: http://www.cs.ox.ac.uk/ian.horrocks/Publications/download/2007/MoHS07a.pdf


Regards,
Ratnesh


From: Peter.Hendler@kp.org<mailto:Peter.Hendler@kp.org> [mailto:Peter.Hendler@kp.org]
Sent: 17 September 2012 22:20
To: Sahay, Ratnesh
Cc: eric@w3.org<mailto:eric@w3.org>; Deus, Helena; kerstin.l.forsberg@gmail.com<mailto:kerstin.l.forsberg@gmail.com>; LINMD.SIMON@mcrf.mfldclin.edu<mailto:LINMD.SIMON@mcrf.mfldclin.edu>; meadch@mail.nih.gov<mailto:meadch@mail.nih.gov>; mscottmarshall@gmail.com<mailto:mscottmarshall@gmail.com>; public-semweb-lifesci@w3.org<mailto:public-semweb-lifesci@w3.org>
Subject: Our white paper on Semantic Node Labeling in Clinical models

This very clearly (to me at least) explains the problem and suggested solution to some of the problems we (Kaiser) see in clinical models.

Wonder who agrees?






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.


________________________________

AMC Disclaimer : http://www.amc.nl/disclaimer


________________________________