The master thesis of Roland Hedayat could also be of interest in this context, see http://www.diva-portal.org/smash/get/diva2:310787/fulltext01.pdf
Copy of abstract text below. Best regards, Erik Sundvall Abstract Semantic Web Technologies in the Quest for Compatible Distributed Health Records Roland Hedayat There is a proliferation of patient bound Electronic Health Record (EHR) data in systems that are incompatible - challenging the goal of granting authorized access to the accumulated medical history of a patient, whenever requested, and whichever the source, in order to secure a safe treatment. A common semantic representation is a prerequisite for validating the semantics of one EHR system against another. Therefore, assessing the semantic compatibility between systems implies having a formal method for extracting their semantics, and for validating the consistency of their combined semantics. A guiding hypothesis is that Semantic Web Technologies and Ontology Web Language (OWL) are potential bridging technologies between the EHRs and medical terminologies, and can be used to represent the combined semantics of the systems to be integrated. Furthermore, that automatic reasoners can perform semantic validation of the combined subsystems. Some experimental steps in this direction are taken, preceded by a discussion on Medical Terminologies, Ontologies, EHR-systems and their interrelationships, and a summary overview of Description Logics, the Semantic Web and the Web Ontology Language, OWL. The OpenEHR reference model is transformed from an XML-schema representation to OWL, and a couple of archetypes are transformed into OWL in a manual procedure. Subsequently, validation runs with a formal reasoner on the transformed results were performed, demonstrating the feasibility of the process. The problems of EHR semantic interoperability are complex. Awareness of the necessity of applying formal semantic methods when dealing with inherently semantic problems will catalyze the process of solving them. Handledare: Daniel Karlsson Ämnesgranskare: Hans Åhlfeldt Examinator: Anders Jansson IT 10 007 onsdag 13 juli 2016 skrev Matheus Wichman <[email protected]>: > I uploaded my ontologies to a Git repository, you can download here > https://bitbucket.org/uhospital/openehr-rm. > > Yes, my approach to handle generics was to create a specialized version > for every possible T type. For example, to create a DV_INTERVAL locked to > DV_QUANTITY I would subclass DV_INTERVAL adding a restriction so lower and > upper properties only accepts DV_QUANTITY values. > > My other problem was with lists. The OWL language doesn't have a > construction for data containers, only RDF do (through rdf:list). I tried > to use a linked list like described here > <http://protege.stanford.edu/conference/2006/submissions/abstracts/7.1_Drummond_listsInProtegeOWL.pdf> > but > I would not be able to create inferences. Currently, lists are represented > just with multiples values for a single property. > > 2016-07-13 4:03 GMT-03:00 Seref Arikan <[email protected] > <javascript:_e(%7B%7D,'cvml','[email protected]');>>: > >> Hi Matheus, >> >> I'd be interested to hear about your problems. I'd also be curious to >> learn how you handled generics/parameterized types. Unless one defines a >> meta layer in owl with semantics for generics, I guess the only option is >> to materialize every possible type parameter T to its own type. >> >> All the best >> Seref >> >> >> On Tue, Jul 12, 2016 at 7:26 PM, Matheus Wichman < >> [email protected] >> <javascript:_e(%7B%7D,'cvml','[email protected]');>> wrote: >> >>> Hi Seref, >>> >>> The ontologies available on the link you posted were developed using an >>> old specification of RM. >>> >>> I converted the current RM to OWL, like Isabel did. However I had some >>> problems to represent list structures. >>> >>> 2016-07-12 13:02 GMT-03:00 Seref Arikan < >>> [email protected] >>> <javascript:_e(%7B%7D,'cvml','[email protected]');>>: >>> >>>> Greetings, >>>> >>>> I was wondering if there is any projects out there that provide an OWL >>>> version of the RM. >>>> >>>> I found this page: http://trajano.us.es/~isabel/EHR/ where an OWL >>>> version is kindly provided, though I'll need to get in touch with the >>>> author to clarify the terms and conditions of use. >>>> >>>> There are papers out there that use an OWL representation of archetypes >>>> etc but I could not find any RM representation other than the link above. >>>> Any pointers would be appreciated. >>>> >>>> Cheers >>>> Seref >>>> >>>> _______________________________________________ >>>> openEHR-technical mailing list >>>> [email protected] >>>> <javascript:_e(%7B%7D,'cvml','[email protected]');> >>>> >>>> http://lists.openehr.org/mailman/listinfo/openehr-technical_lists.openehr.org >>>> >>> >>> >>> >>> -- >>> Atenciosamente, >>> >>> Matheus Wichman >>> >>> _______________________________________________ >>> openEHR-technical mailing list >>> [email protected] >>> <javascript:_e(%7B%7D,'cvml','[email protected]');> >>> >>> http://lists.openehr.org/mailman/listinfo/openehr-technical_lists.openehr.org >>> >> >> >> _______________________________________________ >> openEHR-technical mailing list >> [email protected] >> <javascript:_e(%7B%7D,'cvml','[email protected]');> >> >> http://lists.openehr.org/mailman/listinfo/openehr-technical_lists.openehr.org >> > > > > -- > Atenciosamente, > > Matheus Wichman > -- Sent from mobile.
_______________________________________________ openEHR-technical mailing list [email protected] http://lists.openehr.org/mailman/listinfo/openehr-technical_lists.openehr.org
