Medical applications of enhanced rule-based expert systems

Area of Science:

• Artificial Intelligence
• Medical Informatics
• Expert Systems

Background:

• Classical rule-based diagnostic expert systems face limitations in efficiency and scalability.
• There is a need for advanced techniques to improve the performance of diagnostic systems.

Purpose of the Study:

• To describe efficiency enhancements for classical rule-based diagnostic expert systems.
• To explore the application of these enhancements in medical decision-making tasks.

Main Methods:

• Implementation of a blackboard control structure for parallel knowledge base exploration.
• Integration of taxonomy structures for rapid attention focusing and inheritance-based inference.
• Utilizing the extended FEL-EXPERT shell for case studies.

Main Results:

• The blackboard structure allows parallel processing of multiple knowledge bases.
• Taxonomy structures enable efficient information retrieval and inference through inheritance.
• Four case studies demonstrated the applicability in medical decision-making.

Conclusions:

• The described techniques significantly enhance the efficiency of diagnostic expert systems.
• These enhancements represent a step towards more sophisticated multi-agent decision support tools.
• The findings are applicable to various medical decision-making scenarios.