Just a remark. With the presence of things like MedBiquitous, UMLS, HL7, etc. etc. lurking around, each of them claiming to speak for God, I think that the wisest course for open socerers in medicine to follow is to develop a means of facilely moving between all these candidate data representations. I wonder if machine translation software (for example www.softissimo.com which is truly unbelievable) could help. The idea would be not quite what I think Andrew is proposing. It would be incremental translation between all the candidates.
For another application, namely patient identification, CORBAmed PIDS already does this, but that's integration at the application level, perhaps not what is needed here. And it's for a very small namespace: one person at a time.
The OMG Healthcare Domain Taskforce (aka CORBAmed) has a service called LQS (preferred as Terminology Query Service) that is trying to provide a standard interface to provide equivalences between terms in different namespaces. Tom Culpepper showed me it running on top of UMLS, and, in fact, I think it is fairly simple to put an LQS/TQS wrapper on UMLS. A lot of different terminologies can then be made to coexist, although someone has to do the work of figuring out what SNOMED term maps to what LOINC term, for example. There is no publishing interface for LQS, although some have extended it to enable it to do that, and that would be a logical extension of the interface.
At AMIA2000, there was a paper on Terminology Query Language (TQL) as a dialect of XML so that one could look up terms and their equivalences. It too was running on top of UMLS.
So I agree we need to have a mechanism that is basically a Thesaurus for registering equivalent terms. I would hope that is what Medbiquitous would be pursuing.
Dave
In sum, one might be well-advised *not* to try to develop yet another terminology, but rather develop something that translates between *any* two terminologies...with a minimum of effort.
John
At 03:59 PM 5/29/01, you wrote:
Hello Wayne,
While on the surface, I agree with you, I personally haven't seen much of
substance come from the medical professional societies on collaboration.
That is one of the reasons that I think that individual clinicians working a
few minutes here or there to contribute to a growing OpenOntology might be a
better solution.
I don't really know what is going to work, but I am thinking the smaller and
more international the better.
My .02 euros
Todd Smith <[EMAIL PROTECTED]>
-----Original Message-----
From: Wayne Wilson [mailto:[EMAIL PROTECTED]]
>My 2 cents are to try involve the professional societies, at
>least here in the US, I don't know about elsewhere.
Computer and Computational Sciences
Los Alamos National Laboratory
505-665-1907
