Publications

Relational Storage Techniques Applied to a Medical Expert System

04/24/1984

TIMOTHY KOSCHMANN¹, DAN SOLOMON², FRANK NAEYMI-RAD¹, MARTHA EVENS², MAX HARRY WElL¹, ERIC C. RACKOW¹

¹Department of Medicine, and the Program on Medical Decision Systems, University of Health Sciences/The Chicago Medical School, North Chicago, Ill. 60064
²Illinois Institute of Technology, Dept. of Computer Science, Chicago, Ill. 60616

Abstract
An existing pattern recognition system has been reimplemented, such that the medical know-ledge base is stored using a commercial, rela-tional DBMS. The system, MEDAS, is a general purpose computer-based diagnostic consultant. MEDAS has already been used to create expert systems in the domains of emergency medicine and flight-related disorders. The relational version of MEDAS will be known as MEDAS-R. m is paper will give an overview of the architecture of the complete MEDAS system and will describe the logical structure of a data base for storing medical diagnostic information. Also described, will be the knowledge engineering programs (either implemented or planned), which will be used to create and maintain MEDAS knowledge bases.

Download this publication as a PDF file