An expert system for the identification of meaningful regionswithin MR brain images

摘要

Discusses the design criteria for building a knowledge-basedexpert system that will reliably and efficiently identify normal andabnormal anatomical or physiological regions within medical images. Itsknowledge system architecture is based on the blackboard model and thereasons for choosing this model are discussed. The main features of sucha system have been applied to a prototype expert system for identifyingmeaningful regions within transaxial nuclear magnetic resonance (NMR)brain images