Automated medical student-a computational model of skill acquisition and expert performance

A description of a medical expert system called the automated medical student (AMS) which attempts to combine models of learning with models of expert performance is presented. AMS´s domain is the pediatric subspecialty of neonatology. Unlike traditional expert systems, which are built upon a knowledge base developed through some method of protocol analysis, AMS´s knowledge base is acquired gradually through experience. AMS utilizes several expert systems and artificial intelligence methodologies, including rule- and frame-based reasoning, reasoning from first principles, and a connectionist network model of expert performance