In a reply wrt "On Information and Interoperability" I have noted that
there is a move underway to try & produce an HL7 model (via EMF/MOF) for
use in our /OHT eclipse tooling.
Has anyone looked at an AOM/MOF mapping?
If so any thoughts?
E.g. were one to want to sit down & do some Eclipse OpenEHR tooling then
an obvious contender would be the Eclipse EMF/GMF & that would require a
AOM<>EMF mapping & given EMF is a subset of MOF then ....etc.
Adam
Hi Adam
This is something we would very much like to do. I would propose the following senario:
Develop a template for CCR
Document it (html) and enable data entry
Transform the template to MOF
Create data against the MOF
Transform the data entered against the template to CDA
Compare the data
This would seem useful as a trial.
Cheers, Sam
Adam Flinton wrote:
Sam,
Help me understand this exercise if CCD exists?
Ed
Sam Heard
<sam.heard@oceani
nformatics.com> To
Sent by: adam.flinton@nhs.net, For openEHR
openehr-technical technical discussions
-bounces@openehr. <openehr-technical@openehr.org>
org cc
Subject
04/19/2008 08:26 Re: AOM MOF mapping
AM
Please respond to
For openEHR
technical
discussions
<openehr-technica
l@openehr.org>
Hi Adam
This is something we would very much like to do. I would propose the
following senario:
1. Develop a template for CCR
2. Document it (html) and enable data entry
3. Transform the template to MOF
1. Create data against the MOF
4. Transform the data entered against the template to CDA
5. Compare the data
This would seem useful as a trial.
Cheers, Sam
Adam Flinton wrote:
In a reply wrt "On Information and Interoperability" I have noted
that
there is a move underway to try & produce an HL7 model (via EMF/MOF)
for
use in our /OHT eclipse tooling.
Has anyone looked at an AOM/MOF mapping?
If so any thoughts?
E.g. were one to want to sit down & do some Eclipse OpenEHR tooling
then
an obvious contender would be the Eclipse EMF/GMF & that would
require a
AOM<>EMF mapping & given EMF is a subset of MOF then ....etc.
Adam
Adam & Sam
This is very interesting, kind of relates to my recent post "MDA/MDD & DSL".
For my med student brain I want to clarify that I get what Adam
suggests. MOF has the idea of 4-layer meta-modelling. In the case of
AOM/MOF mapping this would lead to this:
m3 (meta-metamodel) -> MOF
m2 (metamodel) -> AOM and RM (?) [expressed as instance of MOF]
m1 (model) -> Archetypes
m0 (data) -> Archetype instances
Is that correct? I am especially curious whether the Reference Model
(RM) as indicated above also needs to be expressed as MOF in the m2
layer. I would presume so.
As Adam, suggested it makes sense to used the EMF infrastructure &
tools (e.g. have a look at the screen-video
http://redmonk.com/tv/eclipse-emf-demo-large/ ) as their
meta-metamodel Ecore is supposed to be pretty much EMOF (essential
MOF) compliant.
@Sam: If I understand you correctly your trial design starts with an
CCR-openEHR-template (i.e. several aggregated archetypes plus maybe
further constraints). This would be the m1-layer. Before we could do
that we would have to create the generic m2-layer.
Thilo
Hi Ed
The process is really about bringing the clinical specifications into a common framework. From the openEHR perspective this involves:
As we have a growing number of Archetype to CDA transforms this allows production of CDA documents from the openEHR environment in a reusable manner. The full ‘pipeline’ of CCR instance → openEHR → CDA is therefore possible without intervention and with full standardised clinical content validation (as well as any constraints expressed in CCR via the template). openEHR users then have a means of dealing with CCR and CDA documents in the same environment (as well as v2 and XML etc) .
If people are ready to accept such transforms as a wonderful thing (or even useful) and we validate the outputs from the CCD perspective (remember it is a single transform per archetype so it should then work in any CDA document (assuming there is some standardisation in that environment) then it should be possible to get the MOF statement from AOM representation of an archetype. This will require some work but it would reduce concerns in the market.
By the way, what the pipeline offers to vendors and jurisdictions even as it stands is the possibility of building templates (always from archetypes) and creating a template data schema that maps to their own data model. If the data validates, then they can transform their data to openEHR and from there to CDA, CCR, v2 etc without understanding any of the complexities. Of course integration engines will perform something similar on a case by case basis.
I hope that is helpful, Sam
William E Hammond wrote:
Thanks for the response. I am not sure I agree that CCD is a paper, but I
guess time will tell which is the way to go.
Looks like HL7 needs to decide where it fits in today's world and really
promote that position. I for one think CCD has a lot of promise.
Ed
Sam Heard
<sam.heard@oceani
nformatics.com> To
Sent by: For openEHR technical discussions
openehr-technical <openehr-technical@openehr.org>
-bounces@openehr. cc
org
Subject
Re: AOM MOF mapping
04/21/2008 06:41
PM
Please respond to
For openEHR
technical
discussions
<openehr-technica
l@openehr.org>
Hi Ed
The process is really about bringing the clinical specifications into a
common framework. From the openEHR perspective this involves:
links to terminology developments to ensure a sustainable approach
and transformations to a terminology only syntax if that proves
useful
links to implementations of these specifications in openEHR, CEN/ISO
or CDA
CCD is a paper and XML schema exercise to get CCR and CDA into the same
semantic space, but there is no coherent approach as each are XML schemas
and have a lot of attendant paper guides. As openEHR Archetypes are largely
independent of any implementation concerns, it is possible to express the
clinical content of the CCR as a Template entirely in terms of standard
archetypes. From this, a specific schema (Template Data Schema) can be
presented which should ideally map 1:1 with the clinical content of CCR.
This allows integration of CCR into the openEHR space in a controlled
manner with validation via the TDS.
As we have a growing number of Archetype to CDA transforms this allows
production of CDA documents from the openEHR environment in a reusable
manner. The full 'pipeline' of CCR instance -> openEHR -> CDA is therefore
possible without intervention and with full standardised clinical content
validation (as well as any constraints expressed in CCR via the template).
openEHR users then have a means of dealing with CCR and CDA documents in
the same environment (as well as v2 and XML etc) .
If people are ready to accept such transforms as a wonderful thing (or even
useful) and we validate the outputs from the CCD perspective (remember it
is a single transform per archetype so it should then work in any CDA
document (assuming there is some standardisation in that environment) then
it should be possible to get the MOF statement from AOM representation of
an archetype. This will require some work but it would reduce concerns in
the market.
By the way, what the pipeline offers to vendors and jurisdictions even as
it stands is the possibility of building templates (always from archetypes)
and creating a template data schema that maps to their own data model. If
the data validates, then they can transform their data to openEHR and from
there to CDA, CCR, v2 etc without understanding any of the complexities. Of
course integration engines will perform something similar on a case by case
basis.
I hope that is helpful, Sam
William E Hammond wrote:
Sam,
Help me understand this exercise if CCD exists?
Ed
Sam Heard
<sam.heard@oceani
nformatics.com>
To
Sent by: adam.flinton@nhs.net, For
openEHR
openehr-technical technical discussions
-bounces@openehr.
<openehr-technical@openehr.org>
org
cc
Subject
04/19/2008 08:26 Re: AOM MOF mapping
AM
Please respond to
For openEHR
technical
discussions
<openehr-technica
l@openehr.org>
Hi Adam
This is something we would very much like to do. I would propose the
following senario:
1. Develop a template for CCR
2. Document it (html) and enable data entry
3. Transform the template to MOF
1. Create data against the MOF
4. Transform the data entered against the template to CDA
5. Compare the data
This would seem useful as a trial.
Cheers, Sam
Adam Flinton wrote:
In a reply wrt "On Information and Interoperability" I have
noted
that
there is a move underway to try & produce an HL7 model (via
EMF/MOF)
for
use in our /OHT eclipse tooling.
Has anyone looked at an AOM/MOF mapping?
If so any thoughts?
E.g. were one to want to sit down & do some Eclipse OpenEHR
tooling
then
an obvious contender would be the Eclipse EMF/GMF & that would
require a
AOM<>EMF mapping & given EMF is a subset of MOF then ....etc.
Adam
Hi Ed
I am sorry if I sounded disparaging in anyway. I was referring to the
implementation guide which is the basis for various schematron and other
approaches (as I understand it). I am sure that a lot of people will
choose CDA and CCD particularly in the near future. I know you are
interested in the clinical specifications and have formed a clinical
council. I think it would be wonderful to see the European effort line
up with the US effort in the clinical specifications area around the CEN
and hopefully ISO approach. I understand the difficulties.
A small group of enthusiasts working at a distance has got us to this
point. The openEHR Foundation is planning to move into formal
relationships with a number of agencies and we have the prospect of
alignment of a number of initiatives. I believe a small group working on
AOM -> MOF would be very useful and give a way forward to a single
logical representation of clinical content. Clinicians around the globe
will appreciate this.
Cheers, Sam
William E Hammond wrote:
Thanks Sam. I agree with the need for the alignment you suggest. Maybe
the future does hold some hope of working to a common goal.
Ed
Sam Heard
<sam.heard@oceani
nformatics.com> To
Sent by: For openEHR technical discussions
openehr-technical <openehr-technical@openehr.org>
-bounces@openehr. cc
org
Subject
Re: AOM MOF mapping
04/22/2008 01:46
AM
Please respond to
For openEHR
technical
discussions
<openehr-technica
l@openehr.org>
Hi Ed
I am sorry if I sounded disparaging in anyway. I was referring to the
implementation guide which is the basis for various schematron and other
approaches (as I understand it). I am sure that a lot of people will
choose CDA and CCD particularly in the near future. I know you are
interested in the clinical specifications and have formed a clinical
council. I think it would be wonderful to see the European effort line
up with the US effort in the clinical specifications area around the CEN
and hopefully ISO approach. I understand the difficulties.
A small group of enthusiasts working at a distance has got us to this
point. The openEHR Foundation is planning to move into formal
relationships with a number of agencies and we have the prospect of
alignment of a number of initiatives. I believe a small group working on
AOM -> MOF would be very useful and give a way forward to a single
logical representation of clinical content. Clinicians around the globe
will appreciate this.
Cheers, Sam
William E Hammond wrote:
Thanks for the response. I am not sure I agree that CCD is a paper, but
I
guess time will tell which is the way to go.
Looks like HL7 needs to decide where it fits in today's world and really
promote that position. I for one think CCD has a lot of promise.Ed
Sam Heard
<sam.heard@oceani
nformatics.com>
To
Sent by: For openEHR technical discussions
openehr-technical <openehr-technical@openehr.org>
-bounces@openehr.
cc
org
Subject
Re: AOM MOF mapping
04/21/2008 06:41
PM
Please respond to
For openEHR
technical
discussions
<openehr-technica
l@openehr.org>
Hi Ed
The process is really about bringing the clinical specifications into a
common framework. From the openEHR perspective this involves:
links to terminology developments to ensure a sustainable approach
and transformations to a terminology only syntax if that proves
useful
links to implementations of these specifications in openEHR,
CEN/ISO
or CDA
CCD is a paper and XML schema exercise to get CCR and CDA into the same
semantic space, but there is no coherent approach as each are XML schemas
and have a lot of attendant paper guides. As openEHR Archetypes are
largely
independent of any implementation concerns, it is possible to express the
clinical content of the CCR as a Template entirely in terms of standard
archetypes. From this, a specific schema (Template Data Schema) can be
presented which should ideally map 1:1 with the clinical content of CCR.
This allows integration of CCR into the openEHR space in a controlled
manner with validation via the TDS.As we have a growing number of Archetype to CDA transforms this allows
production of CDA documents from the openEHR environment in a reusable
manner. The full 'pipeline' of CCR instance -> openEHR -> CDA is
therefore
possible without intervention and with full standardised clinical content
validation (as well as any constraints expressed in CCR via the
template).
openEHR users then have a means of dealing with CCR and CDA documents in
the same environment (as well as v2 and XML etc) .If people are ready to accept such transforms as a wonderful thing (or
even
useful) and we validate the outputs from the CCD perspective (remember it
is a single transform per archetype so it should then work in any CDA
document (assuming there is some standardisation in that environment)
then
it should be possible to get the MOF statement from AOM representation of
an archetype. This will require some work but it would reduce concerns in
the market.By the way, what the pipeline offers to vendors and jurisdictions even as
it stands is the possibility of building templates (always from
archetypes)
and creating a template data schema that maps to their own data model. If
the data validates, then they can transform their data to openEHR and
from
there to CDA, CCR, v2 etc without understanding any of the complexities.
Of
course integration engines will perform something similar on a case by
case
basis.
I hope that is helpful, Sam
William E Hammond wrote:
Sam,Help me understand this exercise if CCD exists?
Ed
Sam Heard
<sam.heard@oceani
nformatics.com>
To
Sent by: adam.flinton@nhs.net, For
openEHR
openehr-technical technical discussions-bounces@openehr.
<openehr-technical@openehr.org>
org
ccSubject
04/19/2008 08:26 Re: AOM MOF mappingAM
Please respond to
For openEHR
technical
discussions
<openehr-technica
l@openehr.org>
Hi Adam
This is something we would very much like to do. I would propose
the
following senario:
1. Develop a template for CCR
2. Document it (html) and enable data entry
3. Transform the template to MOF
1. Create data against the MOF
4. Transform the data entered against the template to CDA
5. Compare the data
This would seem useful as a trial.Cheers, Sam
Adam Flinton wrote:
In a reply wrt "On Information and Interoperability" I have
noted
that
there is a move underway to try & produce an HL7 model (via
EMF/MOF)
for
use in our /OHT eclipse tooling.Has anyone looked at an AOM/MOF mapping?
If so any thoughts?
E.g. were one to want to sit down & do some Eclipse OpenEHR
tooling
then
an obvious contender would be the Eclipse EMF/GMF & that
would
Has this email gotten through? Wondering since Sam recently described
problems with the list (Eric's SNOMED post).
Is my assumption regarding MOF (see below) right?
Sam
In your opinion what is the advantage of expressing templates in MOF?
Can't the described exersise CCR->openEHR->CDA be done already only
with openEHR/Ocean tools? Is to have a more indepedent intermediary
format?
Thilo
Yes it did. I know; Gmail's refusal of your own emails sent to mailing
lists is a pain in the butt! I had not having mine on my client.
Cheers,
Tim
No Tilo - it is to enable archetypes and templates to be used directly in HL7.
Sam
Thilo Schuler wrote:
Sorry, still don't get it completely.
What do you mean by "use directly in HL7"? Do you refer to HL7 CDA
only or to any HL7 message? Could you please provide a little more
detail how MOF would help with that.
Thanks. Thilo
Thilo Schuler wrote:
Sorry, still don't get it completely.
What do you mean by "use directly in HL7"? Do you refer to HL7 CDA
only or to any HL7 message? Could you please provide a little more
detail how MOF would help with that.Thanks. Thilo
I am looking at it more from the EMF/EMOF direction.
i.e. could one/would one/ how easy would it be to create an EMF/EMOF
model of the AOM which one could then use wrt graphical editing, Model
diffing (e.g. using the EMF compare tooling), Model to model transform etc.
However fitting an OpenEHR artefact within a CDA or similar document
(where both are within the same consistent model) is a worthy goal in
itself.
Adam