> Is this ontology based system similar to GEHR and it's archetype
> definitions?
>
> Thanks,
> --
> Tim Cook, President - FreePM,Inc.
Hi,
Currently, I don't know GEHR enough to be able to compare. I think both
systems are complementary.
I can (quickly) describe what Odyssee is.
LEXIQUE :
The very core of the system is the Lexique : currently a list of more than
35 000 (french) terms, each with a code.
A good description of this set of terms would be "the words of medical
langage" : so, you will find terms of anatomy, pathology, symptoms, units,
biometry...
Example :
Hypertension artérielle / PHTA01
Kg / 2KG001
Poids / VPOID1
and so on... when 2 terms are genuine synonyms, only the last character of
their codes change : HTA / PHTA02
SEMANTIC NETWORK :
Nothing original : links between terms of the Lexique to give sens.
Links can be "is a" or "belongs to" or "has unit"
For example : colitis "is a" inflammatory disease / colitis "belongs to"
colon / weight "has unit" Kg
DESCRIPTION'S TREE :
In Odyssee, we describe all that we can with trees. If we compare the
Lexique with medical vocabulary, trees are sentences made of its words.
Each node of a tree is an object with fields like Lexique's code/Complement
(to store numbers or external codes)/Degree of evidence (from 0=no to
100=certain)
Trees can also contain free text sentences.
So, we can describe a polyp :
Polyp
- size
-- 3 mm
- aspect
-- pediculated
- location
-- left transversal colon
- treatment
-- ablation
--- material
and so on...
But also the whole colonoscopy (the above tree is a subtree of Description,
wich is a subtree of Colonoscopy), or a GP examination, or Biological
results (very narrow tree), or whatever you want.
In Odyssee, each and every structured document is a tree ; you just have to
look at the Lexique term at its root to know what it is.
The whole patient record can even be seen as a huge tree with Lexique's term
Patient as a root.
Trees can be shown *as is* or, for report generation, be translated to
natural langage sentenses (that remains virtual - only tree representation
is stored).
I forgot to say that every Odyssee document can be published as a HTML page,
and when that document is a tree, the tree is hidden in the page.
FILS GUIDES :
Fils guides are probably the major difference between GEHR and Odyssee.
Usually, doctors have first to choose a *model tree* (made by an *expert*)
and instanciate what he wants to describe in it (when I say a *model tree*,
it can be a form as well, a form being a narrow tree).
There are two severe problems with that : first a *model tree* made by an
expert carry expert's habits as well as expert's knowledge (usually
knowledge is close from the leaves), then as you can't mix the trees, you
have to choose that one or that one, but patient can have a proctological
problem and a headache !
It is not easy to describe what Fils guides are, but I will try :
Imagine some expert's trees, then take every branchs and take it apart : you
get a set of *crossings*, each of them with some proposals (say one being
size/aspect/location and another one moderate/mean/severe and so on). Now,
you ask the experts "take each crossing, and tell me what would be the
typicall (and more general) path in which that is a good proposal". The path
is made with terms of the Lexique, and Jokers (* for every sequence, ? for a
single term) for example "colonoscopy/*/polyp" is a path.
Those objects, with a path and a set of proposals, are Fils guides ; for
example (colonoscopy/*/polyp)->size/aspect/location
Now, how does it work ?
When you want to describe something (then building a tree), you can - at
each step - validate a proposal (made by a Fil guide), or pick any term in
the Lexique, or typing free text. If you have completed the tree with a term
of the Lexique (first two choices), the system looks at the path you are now
in, for example, you have just chosen "Dysphagia" and your path now is "GP
examination/reason of encounter/dysphagia" and looks for the Fil guide which
path is the closest from your actual path.
If there is a Fil guide whose path is "GP examination/reason of
encounter/dysphagia", it is immediately chosen, and its proposals become the
new way to go forward. If there is no such path, the system will look for
"GP examination/reason of encounter/X" where dysphagia "is a" X, and so on -
the first (closest) found is chosen.
You can imagine you will find a Fil guide whose path is "GP
examination/*/dysphagia" or even "*/dysphagia" which is even less specific.
That way you can mix experts knowledge, and have it come only when it is
appropriate. You can make a set of Fils guides to a colleague, let him try
it and keep it or put it apart.
It is very important to understand that it is a *non deterministic*
paradigm, that's to say that a very Fil guide can be called from a lot of
different positions in a tree, and if you validate one of its proposals, you
cannot guess which next Fil guide will be elected by the system (since it
depends on the actual user's path).
To end that (very short but still long) description, I must say that a Fil
guide can open a dialog box or attach to dialog box controls. Then,
currently two kind of interfaces exist : genuine trees (tree window control)
or dialog boxes, or a mix of both.
Well, if you are near Paris, I can show.
Philippe
Odyssee project
www.nautilus-info.com
