Medical knowledge representation

Medical knowledge representation is the field that studies how clinical knowledge — diagnostic criteria, treatment protocols, drug interactions, anatomical relationships — is encoded into formal structures that can be processed by computational systems. The field sits at the intersection of medicine, computer science, and cognitive science, and it underlies every clinical decision support system, electronic health record, and diagnostic algorithm in use today. But the act of representation is not neutral. Every formalization is a choice about what to preserve and what to discard, and those choices encode assumptions about what medicine is, what a disease is, and what a patient is.

The dominant paradigm in medical knowledge representation is the ontology approach: building formal taxonomies of diseases, symptoms, and treatments that can be reasoned over by logical inference engines. The SNOMED CT and ICD-10 coding systems are the most widely deployed examples, and they have enabled extraordinary advances in data interoperability, epidemiological tracking, and billing automation. But the ontology approach has a fatal limitation: it treats medical knowledge as static, universal, and context-independent. A diagnosis in ICD-10 is a code, not a story. It captures what the patient has but not who the patient is, how they live, or what they value. The formalization that makes the knowledge computable also makes it inhuman.

The alternative paradigm — the narrative or case-based approach — preserves the contextual richness of clinical reasoning but sacrifices the interoperability that makes the ontology approach powerful. The tension between these two paradigms is not a technical problem. It is a philosophical problem about the nature of medical knowledge itself: is medicine a science of generalizable laws, or a practice of situated judgment? The answer, of course, is both. But no representation system has yet managed to be both at once, and the choice of which paradigm to adopt is a choice about what kind of medicine to practice.