You write:
>And simply sending messages is difficult enough.
>How many ways are there to write dates, names, adressess, etc.
>Before we can use any system, EDIFACT, Internet protocol, SGML, WE must
>decide how to exchange such simple things in a standard way. We must
>provide a standard context (and representation of it) we agree about.
How one choses to write these things is not as important as an INDIRECT
standard for their interpretation. We in America are going to go right
on writing our dates MM/DD/YYYY even though everyone in Europe uses
DD/MM/YYYY... but if one knows that, then the parse is simple. The British
will continue to drive on the "other" side of the road, and we will
continue to use Farenheit and feet and inches (and the Brits "stone?").
If one knows by tagging or labeling what has been done, it doesn't take
much artificial intelligence to calculate an equivalent.
>Then we must unite about some simple medical things/concepts like:
>diagnosis, complaint, symptom, finding, episode, treatmentplan, lab result,
>a bed, a assistent, a doctor, etc, etc
Here again, practice more or less defines a standard meta-outline, from
which varients convenient to specific circumstances can be derived, and
such differences can be managed, if they are derived in an interpretable
way.
>And then we must find a suitable vehiculum to exchange those concepts.
>That vehiculum must be very flexible, forgiving, adaptable, durable,
>usable, etct, etc.
The key criteria exactly, plus openly readable, and requiring minimal
overall housekeeping
>( I happen to think I know what that will be, but any other program,
>standard will be fine with me)
>Greetings
Same
><< end of forwarded material >>
>Gerard Freriks,huisarts, MD
Tom Lincoln
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|