Knowledge representation issues in ontology-based clinical Knowledge Management systems

Abstract

Knowledge Management in healthcare covers a number of diverse practice activity areas that range from admission and accounting to preventive health programmes. From among these areas, clinical knowledge management represents a specific category that poses differentiated problems and requires specific management support. Clinical knowledge as practiced today mixes formally assessed scientific knowledge with a person-culture in which the expertise of the clinician is the key element. When considering standard Knowledge Management life cycles, this entails that the required processes of creation, assessment and dissemination of clinical knowledge assets diverge from other kind of activities in how different kinds of knowledge are handled. Further, the Information Technology support required for clinical knowledge assets is complex and multi-perspective, thus requiring schemas that integrate formally gathered evidence and subjective practical knowledge. This paper Sicilia deals with those differences from the viewpoint of formal ontology as a tool to model the specificities of clinical knowledge. An epistemological account of such knowledge is first provided, which serves to delineate how clinical processes and clinical knowledge management could be aligned. The problems of claim evaluation and representation are then approached from that framework, resulting in a realistic integrated set of design guidelines for clinical knowledge management prepared for use in ontology-based Information Systems.

Towards an ontology for psychological disorders

“…This movement can be extended to medical domains. In this way, modern formal ontology facilitates the creation of knowledge-based systems for managing medical information (Sicilia et al, 2009). Moreover, ontologies constitute an important enhancement in the field, since they allow a better representation of biomedical data, enabling more effective queries, statistical analysis and semantic web searching (Viti et al, 2011).…”

Metadata and Semantics

“…The diagnostic process has traditionally resisted many efforts to be fully automatized [28,35]; it is important to point out that the aim of the NEUROWEB infrastructure is not to replace the clinical expert in the diagnostic activity, but rather to reconstruct the patients' clinical phenotypes, according to the evidences collected -and interpreted -by the medical experts during the diagnostic process; the rationale of using explicit and analytical phenotype formulations is to integrate data from different sources preserving methodological coherence.…”

Ontological modeling at a domain interface: bridging clinical and biomolecular knowledge

“…General-purpose medical ontologies (OBO; The Gene Ontology Consortium; The National Center for Biomedical Ontologies; Galen; PATO; SNOMED-CT) could not be used to meet that end, as they proved unsuitable to represent the specific expert knowledge of the NEUROWEB neurovascular communities (Bard et al ., 2004; Bodenreider et al ., 2007). The diagnostic process has traditionally resisted many efforts to be fully automatized (Miller, 1994; Sicilia et al ., 2009); it is important to point out that the aim of the NEUROWEB infrastructure is not to replace the clinical expert in the diagnostic activity, but rather to reconstruct the patients’ clinical phenotypes, according to the evidences collected—and interpreted—by the medical experts during the diagnostic process; the rationale of using explicit and analytical phenotype formulations is to integrate data from different sources preserving methodological coherence.…”