Re: data modeling multi axis context

You might see what Johns Hopkins is doing and what they are willing to share. They have automated triage and preliminary diagnostic applications, among other things. IF you really need some help, I am available for consulting. I have over 20 years of experience in decision support and data warehouse systems.

Steve

"Scott Buck" <jetta19902001_at_yahoo.com> wrote in message news:96ffb6d7.0305051540.74cd612f_at_posting.google.com...
> Hello -
>
> I am a Java/XML guy and have been given this complex DB project
> that I am not sure how to tackle.
>
> The project involves the modeling of medical concepts their retrieval
> and editing for structured reporting. The basic use case is this -
> when a doctor looks at an X-ray he can point to a spot and then
> describe using a dropdown menu (or some other tree representing
> widget) what he has found. The problem is that it is desired that
> the data model for this widget to be among other things 1) context
> aware (i.e. not reason to show you the brain if the system knows it is
> an x-ray of your toe)
> 2) be able to go down different trees (for example system or
> structure - as in getting to your heart via the cardio-vascular system
> or via the upper body) 3) Easily lend itself to data-mining
> operations such as for exam x which diseases were most commonly found.
> 4) speed is critical
>
> I would seem to me that when designing their system companies like
> yahoo would have the same problem.
>
> What are some good models to store such data?
> What is the good balance between loading the data into memory/ getting
> it from the DB?
> Can you point me to some good sites/ books?
>
> Thank you.
Received on Tue May 06 2003 - 18:27:02 CEST