Problems in the Design of Knowledge Bases for Medical Consultation (original) (raw)

The paper discusses the challenges in designing knowledge bases for computer consultation systems in medicine, particularly focusing on the CASNET/Glaucoma system. It addresses the complexities arising from the heterogeneity of medical knowledge and expert judgment variations. Through the analysis of specific cases and collaboration with ophthalmological clinicians, the paper highlights the successes and expert-level performance achieved in diagnosing and recommending treatment for glaucoma cases, while also pointing out the need for improved structures and functions in knowledge bases to support clinical decision-making.