MYCIN II: design and implementation of a therapy reference with complex content-based indexing

Area of Science:

• Medical Informatics
• Clinical Decision Support Systems

Background:

• Current digital reference search tools (e.g., textword-based) have limitations in clinical practice.
• There is a significant need for improved information retrieval in clinical therapeutics.

Purpose of the Study:

• To develop a prototype system (MYCIN II) for content-based markup and search of a clinical therapeutics textbook.
• To supplement existing keyword searching with a concept-based approach for enhanced precision and recall.
• To address the need for reliable answers to specific clinical questions, such as treatment options for infections.

Main Methods:

• Developed a concept-based markup scheme derived from analyzing clinical questions.
• Applied HTML markup to highlight relevant passages within three textbook chapters.
• Constructed a web-based search interface for the system.
• Integrated semi-automated PubMed querying guided by concept markup and user interaction.

Main Results:

• Successfully created MYCIN II, a functional prototype for content-based search.
• Demonstrated the system's ability to facilitate specific, concept-driven queries.
• Enabled semi-automated querying of PubMed, enhancing research capabilities.

Conclusions:

• MYCIN II offers a promising approach to improve information access in clinical therapeutics.
• Content-based markup and search can overcome limitations of generic search tools.
• The system has the potential to enhance clinical decision-making in infectious disease management.