# Implementation Questions **Category:** [Technical (archive)](https://discourse.openehr.org/c/technical-archive/156) **Created:** 2006-09-07 16:18 UTC **Views:** 3 **Replies:** 5 **URL:** https://discourse.openehr.org/t/implementation-questions/14554 --- ## Post #1 by @ANASTASIOU_A1 Dear All My name is Athanasios Anastasiou and i am with the Signal Processing and Multimedia Communications Research Group at the University of Plymouth, UK\. The university is currently leading a 31 institutions partnership in a 4 year European FP6 program called BIOPATTERN\. More information can be found in http://www.biopattern.org or our revised website at http://www.biopattern.org/Test1 by following the link, "About the NoE\." I have been reading the rather detailed specs documents of openEHR and i believe that i am ready to take the plunge and actually implement a proof\-of\-concept software making use of it\. I am focusing on implementing a subset of the standard and before i start i thought about posting some general questions that i have about openEHR\. 1\. Is there a set of guidelines regarding the implementation technologies used that make a piece of software "openEHR compliant?" For example, the standard specifies that an openEHR repository should be tracking the changes in the records\. Is there a preffered implementation for this or can i use whatever does the job? 2\. The messaging section is blank \(or so i think :\-\) \)\. So i am wondering if there is a messaging pre\-standard or other specification about the exchange of messages specific to openEHR\. If there is no such a thing, should i rather assume that the corbamed or CEN standards cover this section? 3\. What exactly does the ehr\_extract entity represents in everyday medical practice? For example, does it refer to the EHR view that a specialist defines? Does it represent the return results of a query towards the repository? 4\. Just to clarify something\. As far as i understand it, binary files from various modalities are stored as URIs to the underlying file system \(whatever that is\)\. Is this correct? Finally, i would just like to point out that several links, over at www\.openehr\.org that are supposed to be pointing to implementations or implementation details are inoperable or incomplete for example: http://www.shef.ac.uk/uni/projects/bstd/homepa.htm http://www.sincromed.ro/bstd.htm GEHR Australia, several interesting sections in many documents are empty http://www.sdhis.nhs.uk/erdip/public/index.shtml http://www.ehr.chime.ucl.ac.uk/docs/Kalra,%20Dipak%20%28PhD%202002%29.pdf http://www.openehr.org/projects/downloads/EU-EHR-demonstrator-sites.jpg Looking forward to hearing from you Athanasios Anastasiou BIOPATTERN: Data Acquisition Task Force University of Plymouth UK --- ## Post #2 by @Dipak_Kalra Dear Athanasios, I won't reply on your detailed questions \(that for implementer experts\!\), but I can give you a revised link to my Thesis, with apologies: http://eprints.ucl.ac.uk/archive/00001584/ I am delighted that your review of openEHR has left you with a positive desire to implement it, and I hope that our family can prove supportive to you \(and that you will share your progress with us\!\) With best wishes, Dipak --- ## Post #3 by @ANASTASIOU_A1 Dear Dipak Thank you for your email and the revised link\. My initial intention was to post this email to the implementers mailing list \(and that is why i left the subject as it is\) but when i reviewed the content of my message i thought that it was more general technical questions than specific implementation issues\. I will certainly share my progress through this mailing list as i will be "treading" along\. Possibly by more questions :\-\) All the best Athanasios Anastasiou BIOPATTERN:Data Acquisition Task Force University of Plymouth UK Dipak Kalra wrote: --- ## Post #4 by @thomas.beale Athanasios Anastasiou wrote: > Dear All > > My name is Athanasios Anastasiou and i am with the Signal Processing and > Multimedia Communications Research Group at the University of Plymouth, UK\. > > The university is currently leading a 31 institutions partnership in a 4 > year European FP6 program called BIOPATTERN\. More information can be > found in http://www.biopattern.org or our revised website at > http://www.biopattern.org/Test1 by following the link, "About the NoE\." > > I have been reading the rather detailed specs documents of openEHR and i > believe that i am ready to take the plunge and actually implement a > proof\-of\-concept software making use of it\. > > I am focusing on implementing a subset of the standard and before i > start i thought about posting some general questions that i have about > openEHR\. >   Hi Athanasios, my first question is: what technology do you want to to work in? There is a lot of open source work already done in Java, Eiffel and \.Net for openEHR which you can use\. > 1\. Is there a set of guidelines regarding the implementation > technologies used that make a piece of software "openEHR compliant?" > For example, the standard specifies that an openEHR repository should be > tracking the changes in the records\. Is there a preffered implementation > for this or can i use whatever does the job? >   What you are after is conformance criteria and tests, which have not yet been published\. Conformance, the way we see it, can be tested only on concrete products, and is based on implementation technology specifications, such as the XML schemas, and \(in the future\) service interfaces\. Currently, if your system can read and write data compliant to the published XML\-schema for the reference model, and process ADL 1\.4 archetypes, it would be minimally openEHR\-compliant\. We will develop test frameworks to perform detailed tests as time goes on and resources allow\. > 2\. The messaging section is blank \(or so i think :\-\) \)\. So i am > wondering if there is a messaging pre\-standard or other specification > about the exchange of messages specific to openEHR\. If there is no such > a thing, should i rather assume that the corbamed or CEN standards cover > this section? >   The openEHR EHR Extract specification will be published in its first draft in the next few weeks, and is significantly more powerful than the CEN specification, but contains a direct mapping to it\. If you want a private draft, let me know\. > 3\. What exactly does the ehr\_extract entity represents in everyday > medical practice? For example, does it refer to the EHR view that a > specialist defines? Does it represent the return results of a query > towards the repository? >   Queries can return finer\-grained results than an EHR Extract\. An EHR Extract is designed for the results of queries for parts of health records, between two distinct EHR systems, usually in 2 care delivery enterprises \- the idea is that the receiver system integrates the extract of the EHR it receives into its local copy of the health record for the patient in question; the results will then be visible to local user applications at the receiver end\. > 4\. Just to clarify something\. As far as i understand it, binary files > from various modalities are stored as URIs to the underlying file system > \(whatever that is\)\. Is this correct? >   If you mean things like CAT scan images, X\-ray images and so on, yes, usually, these are stored in dedicated PACS, with links an thumbnails \(or "normal" images for a typical desktop computer\) embedded in the EHR\. There are two reasons to do this\. One is size and performance; secondly, the EHR usually needs the \_report\_ written by the radiologist or other specialist, plus a few diagnosticaly relevant images, whereas the PACS will contain the whole image set \(for some time at least\)\. > Finally, i would just like to point out that several links, over at > www\.openehr\.org that are supposed to be pointing to implementations or > implementation details are inoperable or incomplete for example: >   I have fixed as many of these as I can, but as you might imagine, some things are disappearing due to age\.\.\. \- thomas beale --- ## Post #5 by @ANASTASIOU_A1 Dear Thomas Thank you very much for your response, please see below > my first question is: what technology do you want to to work in? There > is a lot of open source work already done in Java, Eiffel and \.Net for > openEHR which you can use\. The whole system is a GRID application supported by the Globus Toolkit version 4\.0\. Maybe it was not clear in my last email but i was more concerned with the back\-end part\. That is, the technology used to store the EHR\. I am more inclined to use a database rather than individual XML files\. So, the EHR packages \(demographics, composition, folders, etc\) will be handled at the database level and the individual datasets \(imaging, electrophysiology, etc\) will be handled by services of the GRID middleware that unify the storage nodes\. It would be interesting to have a look at the Java and \.NET source but i can not find it in the web part of svn \(at least by looking under the /app directory \)\.\.\.Any further directions on this? >> 2\. The messaging section is blank \(or so i think :\-\) \)\. So i am >> wondering if there is a messaging pre\-standard or other specification >> about the exchange of messages specific to openEHR\. If there is no such >> a thing, should i rather assume that the corbamed or CEN standards cover >> this section? >>   > > The openEHR EHR Extract specification will be published in its first > draft in the next few weeks, and is significantly more powerful than the > CEN specification, but contains a direct mapping to it\. If you want a > private draft, let me know\. Yes please, that would be great\. This email address can be used for this, however if it is a very big file \(>5MB\) please let me know by a private email\. I would like to thank you for your answers in my questions 1,3,4 of the previous email\. This was all very clear\. All the best Athanasios Anastasiou BIOPATTERN:Data Acquisition Task Force University of Plymouth UK --- ## Post #6 by @thomas.beale Athanasios Anastasiou wrote: > Dear Thomas > > Thank you very much for your response, please see below > >> my first question is: what technology do you want to to work in? There >> is a lot of open source work already done in Java, Eiffel and \.Net for >> openEHR which you can use\. >>     > > The whole system is a GRID application supported by the Globus Toolkit > version 4\.0\. Maybe it was not clear in my last email but i was more > concerned with the back\-end part\. That is, the technology used to store > the EHR\. I am more inclined to use a database rather than individual XML > files\. So, the EHR packages \(demographics, composition, folders, etc\) > will be handled at the database level and the individual datasets > \(imaging, electrophysiology, etc\) will be handled by services of the > GRID middleware that unify the storage nodes\. > > It would be interesting to have a look at the Java and \.NET source but i > can not find it in the web part of svn \(at least by looking under the > /app directory \)\.\.\.Any further directions on this? >   For the Java source, it is best to talk to the Java developers and/or join the Java project mailing list \(see http://svn.openehr.org/ref_impl_java/TRUNK/project_page.htm for links\)\. This implementation uses MySQL as the back\-end\. Ocean Informatics makes its \.Net back\-end \(SQL\-Server 2005\-based\) available for academic use\. Please email privately with more details regarding this\. \- thomas beale --- **Canonical:** https://discourse.openehr.org/t/implementation-questions/14554 **Original content:** https://discourse.openehr.org/t/implementation-questions/14554