The management of ES and DBMS: case of the pancreatic diseases

The authors provide a brief background on the various types of interaction between database systems and expert systems (ES). They then discuss the difficulty in comparing data about pancreatic cancer. They attribute this difficulty to three causes: (1) lack of homogeneity in data collecting and minimum standardization; (2) extreme variability in diagnostic evaluation, because of different tools and data available and procedures used; and (3) prevailing subjectivity of the surgeon in the prognostic-classifying phase. A system architecture using a combined database/expert-system approach is proposed.<>