Thank you Heather for the Link and Congrats for the new born website!
As far as I remember form the tutorial session, you-Heather- started creating an archetype by a brain storming of what any clinician may ask about smoking or what She may examine during the caring process of a smoker patient,…, (I’m not sure about the kind of disease we discussed that day but surely you remember that example)
is it what you-or any other researchers in openEHR area- wanted to show us or I just understood it the way I am used to!
Actually, I’m really stuck into this kind of thinking about Archetypes and Templates and maybe it’s because of what we have done up to now in our project creating forms for Clinicians.
Briefly, I can categorize what I have in my mind as the methodology of creating an application based on openEHR as bellow:
1- organize number of questions your have in your mind regarding this disease
2- Analyze questions to find if there is any scheme for Archetype or Template in them
3- Search for existing Archetypes that may fairly act as the basis for templates you need for this case
4- Create new Archetypes if needed.
5- Combine Archetypes and create proper templates
6- use templates in Application!
Cheers
-Paria
PhD Student
IDC | Interaction Design Collegium
Department of Computing Science and Engineering
Chalmers University of Technology
Email: hajar.kashfi@chalmers.se
Office:+46 (0)31 7725407
Mobile Phone: +46 (0)707222815
Postal adress:
IT University of Göteborg
412 96 Göteborg, Sweden
Visit: Room Simula B, House Svea, Campus Lindholmen
archetypes model what clinicians would want to record about one
discrete clinical idea/concepts. This is done in a maximum dataset
fashion (meaning a single clincician would not need certain parts of
the archetype) and forms the basis for semantic interoperability. Such
archetypes are reused in many different use-cases. So the 'tobacco
use' archetype does NOT aim at a certain disease. It should be used
e.g. in a template of a lung cancer clinic and in a GP check-up
template. This is why the maximum dataset is so important -> the
archetype should include what the specialist and what the GP want to
record. In the use-case specific template unnecessary archetype parts
can be "stripped out".
What we did in the MIE session was to start the process of creating
an archetype to record a patients tobacco use. Clinicians decide what
goes in it (i.e. what they would normally as and note in narrative).
Thus it contains statements about relevant past events like 'age
commenced' and current facts e.g. 'average daily consumption' or the
'readiness to change the habit' (actually 'change' is another discrete
concept and therefore modelled as a reference to another archetype).
Your approach mentioned below seems reasonable. Step 1 would be the
brainstorming (we often do this using mind-mapping software). As you
correctly mentioned it is important to use existing archetypes if
possible. This modularity and reusability is a strength of openEHR is
what we think the main key for semantic interoperability (reusability
on a large scale is only possible through the maximum dataset
approach!).
I know that mind mapping view is available for the archetypes on the openehr site. Ok!
I have always wondered about the xml schema used for the conversion from adl to mind map.
Since mind maps are useful in capturing clinicians' output during a facilitation session. Is there a reversely engineered xml schema that can automatically convert mind map xml(say generated through Freemind) to the adl? If there is, I will like to have a look and if not how can one be developed.
The idea is to abstract away the adl from clinicians and present only the mind maps as representative of the maximum data set of each archetype.
The archetypically minded can they use xml schema and the adl to ensure all kinds of other necessary mappings for different interoperabilty requirements and interests.
It is important to separate the modelling tool for clinicians (eg mind maps) from that for the 'computerati'(adl, schemas etc)
What a clinician group would need to develop new archetypes is the basic knowledge of mind mapping and openehr archetype structures.
Thank you Thilo for your complete answer to my questions.
Hi Päria,
archetypes model what clinicians would want to record about one
discrete clinical idea/concepts. This is done in a maximum dataset
fashion (meaning a single clincician would not need certain parts of
the archetype) and forms the basis for semantic interoperability. Such
archetypes are reused in many different use-cases. So the ‘tobacco
use’ archetype does NOT aim at a certain disease. It should be used
e.g. in a template of a lung cancer clinic and in a GP check-up
template. This is why the maximum dataset is so important → the
archetype should include what the specialist and what the GP want to
record. In the use-case specific template unnecessary archetype parts
can be “stripped out”.
What we did in the MIE session was to start the process of creating
an archetype to record a patients tobacco use. Clinicians decide what
goes in it (i.e. what they would normally as and note in narrative).
Thus it contains statements about relevant past events like ‘age
commenced’ and current facts e.g. ‘average daily consumption’ or the
‘readiness to change the habit’ (actually ‘change’ is another discrete
concept and therefore modelled as a reference to another archetype).
Your approach mentioned below seems reasonable. Step 1 would be the
brainstorming (we often do this using mind-mapping software). As you
correctly mentioned it is important to use existing archetypes if
possible. This modularity and reusability is a strength of openEHR is
what we think the main key for semantic interoperability (reusability
on a large scale is only possible through the maximum dataset
approach!).
Hope this helped.
Cheers, Thilo
Thank you Heather for the Link and Congrats for the new born website!
As far as I remember form the tutorial session, you-Heather- started
creating an archetype by a brain storming of what any clinician may ask
about smoking or what She may examine during the caring process of a smoker
patient,…, (I’m not sure about the kind of disease we discussed that day
but surely you remember that example)
is it what you-or any other researchers in openEHR area- wanted to show us
or I just understood it the way I am used to!
Actually, I’m really stuck into this kind of thinking about Archetypes and
Templates and maybe it’s because of what we have done up to now in our
project creating forms for Clinicians.
Briefly, I can categorize what I have in my mind as the methodology of
creating an application based on openEHR as bellow:
1- organize number of questions your have in your mind regarding this
disease
2- Analyze questions to find if there is any scheme for Archetype or
Template in them
3- Search for existing Archetypes that may fairly act as the basis for
templates you need for this case
4- Create new Archetypes if needed.
5- Combine Archetypes and create proper templates
6- use templates in Application!
Cheers
-Paria
PhD Student
IDC | Interaction Design Collegium
Department of Computing Science and Engineering
Chalmers University of Technology
Email: hajar.kashfi@chalmers.se
Office:+46 (0)31 7725407
Mobile Phone: +46 (0)707222815
Postal adress:
IT University of Göteborg
412 96 Göteborg, Sweden
Visit: Room Simula B, House Svea, Campus Lindholmen
PhD Student
IDC | Interaction Design Collegium
Department of Computing Science and Engineering
Chalmers University of Technology
Email: hajar.kashfi@chalmers.se
Office:+46 (0)31 7725407
Mobile Phone: +46 (0)707222815
Postal adress:
IT University of Göteborg
412 96 Göteborg, Sweden
Visit: Room Simula B, House Svea, Campus Lindholmen
Dear Paria, Thilo, Koray, Thomas and all.
Since Thilo mentions mapping, here is my take...
I know that mind mapping view is available for the archetypes on the openehr site. Ok!
I have always wondered about the xml schema used for the conversion from adl to mind map.
There is none…we go directly from adl to the mindmap, e.g. via the archetype parsed by the Java ADL Parser.
To use XSLT, you’d need to go via the XML archetype; certainly possible but not what we do.
Since mind maps are useful in capturing clinicians' output during a facilitation session. Is there a reversely engineered xml schema that can automatically convert mind map xml(say generated through Freemind) to the adl? If there is, I will like to have a look and if not how can one be developed.
There certainly is none. I doubt this is really feasible - you would need to express a lot of semantics in the mindmap then. You could maybe have a tool that generates some basic ADL from the mindmap…but that’s definitely not round trip then.
We have concentrated on the other way round so far, in the upcoming Knowledge Manager, ADL is use to generate the freemind mindmap on the fly. So you can get the mindmap in an instant.
Cheers
Sebastian
The idea is to abstract away the adl from clinicians and present only the mind maps as representative of the maximum data set of each archetype
The mind map view of the archetypes is a direct rendering by the Java Parser then. I see the point.
Let's look at it like this! Is there an opportunity for this kind of xml-based round tripping? The reason is xml seems to be the standard for achieving any standardised interoperability now and probably in the future.
Could we not use the xml archetypes as you suggested to create a platform for abstracting the adl.
Xml is widely supported:
It may allow integration among the different mapping standards.
What semantics would one have to build into the mind map to make it work?
Does this sound like another area of research?Would it be worthwhile?
Thanks again
Olusegun
Olusegun Odujebe,MD,CISA,PMP
Enthusia Consulting Ltd.
Lagos Nigeria.
...There is a spirit in man and the inspiration of the almighty gives him understanding. Job 32:8
The mind map view of the archetypes is a direct rendering by the Java Parser then. I see the point.
Let's look at it like this! Is there an opportunity for this kind of xml-based round tripping? The reason is xml seems to be the standard for achieving any standardised interoperability now and probably in the future.
Could we not use the xml archetypes as you suggested to create a platform for abstracting the adl.
Xml is widely supported:
It may allow integration among the different mapping standards.
What semantics would one have to build into the mind map to make it work?
Does this sound like another area of research?Would it be worthwhile?
Sure, xml is widely supported and that is one reason why we have xml
archetypes.
So ok to generate a XSLT that transforms into a mindmap format such as
freemind
But in order to do it the other way round, you need all the information
that is in the archetype in the mindmap.
This includes node ids, various languages (of which you'd only want one
usually in a mindmap), detailed info about each item, form cardinality,
occurrences, etc.
All this in a mindmap format, which gives you a lot of freedom, but not
necessary computable expressiveness usually.
As a research topic, why not. I agree it would be nice to be able to
create a mindmap and convert this into an archetype somehow.
Also the XML-archetype-to-mindmap-XSLT is certainly worth doing and may
be a good starting point anyway.
Such a transformation would be a great resource for openEHR in general.
I had been thinking about this as part of the NHS work where it would
be very helpful to do some of the original work in a mindmapping tool
then import it to one of the archetype editors.
I don't think the process needs to be too clever about node ids,
cardinality, typing etc as even just being able to import a simple
structure as text elements/ clusters would save a lot of typing and
drag-dropping. The editor could then be used to refine/refactor
datatypes etc This, of course would be a one-off exercise and not
'round-trip' as has been suggested but would be very helpful as a
start point.
Ian
Dr Ian McNicoll
office / fax +44(0)141 560 4657
mobile +44 (0)775 209 7859
skype ianmcnicoll
Consultant - Ocean Informatics ian.mcnicoll@oceaninformatics.com
Consultant - IRIS GP Accounts
Member of BCS Primary Health Care Specialist Group – www.phcsg.org
this is a bit like you can copy paste from Mindmanager and other Tools
to Word etc and get a structured list of the mindmap.
Limited of course compared to (the difficult or impossible) round-trip,
but still very useful...
Good to hear from you. The XML for the mind map in the tooling environment is based on FreeMind and is very simple. I would like to be able to turn this into archetypes directly as well, although obviously there would be some intermediate steps.
The other work flow that we are facing is screen design and linking this to archetypes (see Paria and paper forms → computable data). This is another sort of brain storming that people want to start from - clearly this will not be a simple mind map.
So we have the possibility of going from free mind to archetypes - perhaps building archetypes in a more MindMap sort of interactive paradigm anyway? And also, designing forms and then linking them to templates and data.
