, 2005, 2

50-


Abstracts

RETROSPECTIVE ANALYSIS OF MEDICAL EXPERT SYSTEMS

B.A. Kobrinskiy

Various decisions in expert systems 10 20-years prescription and later are considered, allowing to raise efficiency of extraction of expert knowledge and representing interest for users. These approaches can be useful at construction of new intellectual systems of medical purpose.

ON SIMPLE AND COMPLEX THINGS IN INFORMATICS

M.A. Shifrin

In the 60-th years of 20 century I.Gelfand and his colleagues introduced the idea that description of live and complex artificial systems requires a variety of languages adequate to the goals of investigators. The paper is focused on use of this idea for understanding different issues of informatics.

USING OF DIRECTED LEARNING METHOD IN MEDICAL MAKING DECISIONS SUPPORTING SYSTEMS

Karp V.P.

For the last 50 years the Russian in medical information science the great improvements were achieved in using newest scientific results in public health service and health protection. Computer systems for supporting medical decisions are now widely adopted. Direct learning method, which was created 30-40 years ago, showed its effectiveness in solving many problems of medical diagnostics.

PROBLEMS OF INFORMATION INTERACTION IN MEDICINE

I.P.Lukashevich

An overall objective of the present work is development of methods of the organization of the medical information which allow to unify representations of knowledge; to establish substantial connections and to reduce volume of a material, having increased, thus, it self-descriptiveness, that should promote increase of efficiency of information interaction in medicine.

ARTIFICIAL NETWORKS AND CONCEPT OF AUTO DIAGNOSTIC SYSTEM OF HUMAN

Chernavsky D.S., Rodshtat I.V., Karp V.P.

The concept of auto diagnostic system is formulated. According this concept there are two systems homeostasis: auto therapeutic (effectors) subsystem and auto diagnostic (informational) one. The structure and functions of the auto diagnostic system are considered in moderns positions of recognition theory, neurophysiology and neurons computing. The theory construction of the optimal recognition neurons computer is discussed. The experimental data (morphological, biochemical and so on) on the structure and function of the Rexed plates are considered. It is shown, that structure andfunctions of recognition neurons computer and Rexed plates are close each other.

ONTOLOGY INTEGRATION AND ELICITATION OF HOLISTIC KNOWLEDGE

Voinov A.V.

A new approach is proposed, which aims at elicitation of holistic knowledge, in particular of one related to intuition and imaginative thinking, which play a major role in the practice of experts in domains of medicine. The approach is based upon a combination of traditional and original methods of psychosemantics (vocabulary scaling), methods of ontology integration and the algebra and the logic of Kripke worlds.