[Gnumed-devel] Re: [wxPython-users] Re: Medical Expert System GUI | Anchoring widgets

[Karsten Hilbert]

On Tue, May 18, 2010 at 02:49:17AM -0700, Dr.James wrote:

> We have a few more doubts - 1) is there a sizer that arranges items
> into rows by itself. For eg.if we have 10 items in a horizontal
> boxsizer it just fits the screen, and when we add two more it goes out
> of screen, is it possible that those new items be added below (in a
> new row)?

If you want it to flow over there's something called similar
to flowsizer but it's not directly in the standard library.

Also, you could use a grid sizer.

> This is important because when we use splitter window and
> some one resizes the window, the widgets remain in their exact
> position, leaving the expanded space empty.

This sounds like you didn't set the proportion properties
properly :-)

> 2) Our expert system currently works under CLI - Command line
> interface and we plan to integrate this facility in GUI too.(Some
> doctors may find it easy to type y/n or 1,2,3 options.) So is there an
> easy way to integrate shell prompt like interface into GUI?

Not that I know.

> > What GNUmed would like to do (at which point we'd want to
> > move further discussion to a more appropriate list such as
> > gnumed-devel which I CCed):
> 
> Sure, we can do that or have a separate group for this project.

I'm fine hosting this discussion on gnumed-devel but would
be fine with it continuing on wxpython-users, too, as long
as subscribers of the latter don't consider this spamming
their list.

> Currently we are two of us who are working on GUI, and it will help a
> lot to have help from more experienced developers. I think we should
> begin by building a standalone application that can interact with
> expert system, and once we do that we can integrate this into GNUmed.

It may be an idea to develop that GUI as a GNUmed plugin.
After all that will gain you fully fledged patient handling.

> Does GNUmed have integrated workflow like feature? For eg. a nurse
> might take vitals and weight and enter into the system, and by the
> time the patient comes in front of physician, his basic demographics
> and vitals are already present.

Sure, GNUmed doesn't care who enters what when. You may need
to be a bit more specific of what you expect from
"integrated workflow".

> If yes, we can further expand this
> into a workflow where the physician assistant interacts with the
> patient, collecting symptoms based on questions asked by expert
> system, feeds them, and by the time the patient arrives to the
> physician a detailed history taking is almost done,

That would work just fine with a plugin in GNUmed. We would
just need to decide on a way of storing data collected. Most
of it would be patient-related data independant of the
expert system (and GUI) anyway and the expert system (or its
frontend) should be taught to retrieve data from the patient
file.

> > Another important area for an expert system would be
> > checking of adherence of patient state (as documented in the
> > database) with known guidelines for the health conditions
> > known to exist in the patient. IOW, monitor management of
> > care.
> 
> Since we want the expert system to formulate the management, it will
> adhere to the guidelines.

Yes, but the expert system should be able to *apply* its
knowledge about guidelines to a specific patient file and be
able to share its findings (say, areas of non-adherence)
back into the EMR.

> > It is rather unlikely doctors will want to enter data in
> > order to get trivial suggestions (such as "- consider
> > pneumonia"). They will welcome "trivial" hints if they
> > *don't* have to enter anything *additionally*, however.
> > OTOH, they would be willing to answer a few questions from
> > the expert system if it helps them detect cases of rare
> > diseases ("- consider Winiwarther-Burger syndrome").
> 
> Well, Dr.Karsten - the situation here (in third world countries) is
> pretty grim - you would be astonished to know about the disastrous
> system here.

True enough - I didn't think of such circumstances. It is of
course useful to make sure typical, frequent diagnoses are
not forgotten because of lack of information or lack of
staff.

> solid in these conditions. Ofcourse, we will want to gradually add
> more and more knowledge to diagnose rare disorders. (This reminds me
> that we also have to discuss the GUI for teaching the system and
> inputing knowledge).

I think you should really look at OpenMRS.

> > There's basically two schools of thought here. The one you
> > put forth and the other one that presents the argument that
> > controlled vocabs *alone* cannot ever capture the whole (or
> > even true) story. Which I tend to agree with from the last
> > 10 years of community medical care.
> 
> I would beg to differ here, the symptoms can vary in many ways, but
> since we NEED to have them formalized (for the sake of expert system),
> i see no problem in solving the research thing.

Surely one should try to reap benefits from expert systems.
If expert systems require controlled vocabularies because
natural language processing is lacking in general - so be
it. One should, however, keep in mind that

a) controlled vocabularies DO NOT *capture* the patient story

        (what they do is *formalize* and/or *normalize* it)

b) for that reason they are NOT first choice for capturing
   medical care

        (but they need to be attached to EMRs in order to formalize
         the narrative on a secondary level such that statistics,
         searching, and processing can produce better results - that's
         what all those coding systems are all about)

Karsten
-- 
GPG key ID E4071346 @ wwwkeys.pgp.net
E167 67FD A291 2BEA 73BD  4537 78B9 A9F9 E407 1346