Authors:
Eriksson Monteiro
;
Pedro Sernadela
;
Sérgio Matos
;
Carlos Costa
and
José Luís Oliveira
Affiliation:
Institute of Electronics and Telematics Engineering of Aveiro (IEETA) and University of Aveiro, Portugal
Keyword(s):
Semantic Web, Healthcare Information Management, Clinical Reports, Radiology, Text-mining.
Related
Ontology
Subjects/Areas/Topics:
Biomedical Engineering
;
Data Engineering
;
Databases and Datawarehousing
;
Enterprise Information Systems
;
Health Information Systems
;
Information Systems Analysis and Specification
;
Knowledge Management
;
Ontologies and the Semantic Web
;
Semantic Interoperability
;
Society, e-Business and e-Government
;
Web Information Systems and Technologies
Abstract:
The tremendous quantity of data stored daily in healthcare institutions demands the development of new methods to summarize and reuse available information in clinical practice. In order to leverage modern healthcare information systems, new strategies must be developed that address challenges such as extraction of relevant information, data redundancy, and the lack of associations within the data. This article proposes a pipeline to overcome these challenges in the context of medical imaging reports, by automatically extracting and linking information, and summarizing natural language reports into an ontology model. Using data from the Physionet MIMIC II database, we created a semantic knowledge base with more than 6.5 millions of triples obtained from a collection of 16,000 radiology reports.