Knowledge-Based Induction of Clinical Prediction Rules

ABsTRAcT This chapter describes how to integrate medical knowledge with purely inductive (data-driven) methods for the creation of clinical prediction rules. It addresses three issues: representation of medical knowledge , secondary analysis of medical data, and evaluation of automatically induced predictive models in the context of existing knowledge. To address the complexity of the domain knowledge, the authors have introduced a semio-fuzzy framework, which has its theoretical foundations in semiotics and fuzzy logic. This integrative framework has been applied to the creation of clinical prediction rules for the diagnosis of obstructive sleep apnea, a serious and under-diagnosed respiratory disorder. The authors use a semio-fuzzy approach (1) to construct a knowledge BLOCKINbase BLOCKINfor BLOCKINthe BLOCKINdefinition BLOCKINof BLOCKINdiagnostic criteria, predictors, and existing prediction rules; (2) to describe and analyze data sets used in the data mining process; BLOCKINand BLOCKIN(3) BLOCKINto BLOCKINinterpret BLOCKINthe BLOCKINinduced BLOCKINmodels BLOCKINin BLOCKINterms BLOCKINof BLOCKINconfirmation, BLOCKINcontradiction, BLOCKINand BLOCKINcontribu-tion to existing knowledge.