Medical Expert Systems
Donna L. Hudson · Wiley Encyclopedia of Biomedical Engineering · 2006
Abstract The first medical expert system was designed over 30 years ago. Since that time, advances in both computer technology and software design have permitted the development of increasingly sophisticated models. Most medical expert systems are designed to function as decision support tools. The original systems used techniques from artificial intelligence as a basis for developing models that were symbolic in nature in which the reasoning structure and knowledge base were separate entities. The hallmark of these systems is the use of natural language, the easy representation of concepts, and the ability to provide explanations. Relevant artificial intelligence methodologies include knowledge representation, knowledge acquisition, and problem solving. In the course of thirty years, many additional techniques have been employed, including causal and deep reasoning, fuzzy logic, approximate reasoning, hybrid systems, consequential reasoning, and intelligent agents. Hybrid systems and intelligent agent models meld symbolic techniques with data‐based models to bring all available information to bear on a problem. Medical applications include all major specialties and range from very specific areas such as antibiotic therapy to entire domains such as internal medicine.