An expert
system can be defined as: a computer program that contains and
uses knowledge to analyze data and to suggest problem solutions or user
actions. The name comes from the “simulation” of
the human expert behavior.
Medicine has been one of the first application
fields for expert systems since ‘70s.
A typical feature of expert
systems is the separation between the knowledge and the program that contains
and manages the knowledge itself.
Shortly, to build and validate a medical expert system, the following
are required:
- The knowledge source – one or more individuals, expert in
a specific medical field. The expert role is close to the author of
a book, with the difference that knowledge in the expert system becomes
"active".
- The
knowledge engineer, who transfers the knowledge from the human expert
to the program. This role requires both medical and informatics expertise.
- The
program that supports the knowledge
representation and makes it active,
called "expert system shell".
- An extended clinical validation
cycle, with significant number of users and cases.
Euristic includes an “expert system shell”, that may be
used for expert systems in different medical areas, each one able to
operate on every data contained in the hospital information system. Components
of Euristic team have been working in knowledge engineering since 1991
for the development of a neurological expert system.
This system is in clinical routine use in a university department of
neurology in the United States with over 22,000 cases. |