Effective domain-dependent reuse in medical knowledge bases

Area of Science:

• Medical Informatics
• Knowledge Engineering
• Artificial Intelligence in Medicine

Background:

• Knowledge reuse is crucial for medical knowledge-based systems.
• Current approaches focus on general-purpose modules, often neglecting medical practice specifics.
• Existing methods may not align with knowledge engineers' domain-specific goals.

Purpose of the Study:

• To propose a pragmatic approach to knowledge reuse in medical systems.
• To introduce the concept of "domain-dependent limited reuse".
• To present practical mechanisms for achieving this type of reuse and variation.

Main Methods:

• Developing a knowledge representation framework combining objects and production rules.
• Proposing three novel mechanisms derived from object-oriented and rule-based programming.
• Evaluating the effectiveness of these mechanisms for limited reuse and adaptation.

Main Results:

• The proposed mechanisms facilitate domain-dependent limited reuse of medical knowledge.
• These mechanisms allow for the introduction of useful, controlled variations in medical expertise.
• The approach is more practical for knowledge engineers than general-purpose reuse.

Conclusions:

• Domain-dependent limited reuse is a viable and effective strategy for medical knowledge systems.
• Combining object-oriented and rule-based programming offers practical solutions for knowledge reuse and adaptation.
• This pragmatic approach better serves the needs of knowledge engineers in specialized medical domains.