An ontological modeling approach to cerebrovascular disease studies: The NEUROWEB case

  • Authors:
  • Gianluca Colombo;Daniele Merico;Giorgio Boncoraglio;Flavio De Paoli;John Ellul;Giuseppe Frisoni;Zoltan Nagy;Aad van der Lugt;István Vassányi;Marco Antoniotti

  • Affiliations:
  • Dipartimento di Informatica, Sistemistica e Comunicazione (DISCo), Universití degli Studi di Milano Bicocca, U14 Viale Sarca 336, I-20126 Milan, Italy;Terrence Donnelly Centre for Cellular and Biomolecular Research (CCBR)/Banting and Best Department of Medical Research, University of Toronto, 160 College Street, Toronto, Ontario, Canada M5S 3E1;Istituto Neurologico Carlo Besta, Via Celoria 11, I-20133 Milan, Italy;Dipartimento di Informatica, Sistemistica e Comunicazione (DISCo), Universití degli Studi di Milano Bicocca, U14 Viale Sarca 336, I-20126 Milan, Italy;Department of Neurology, University of Patras, Greece;Dipartimento di Informatica, Sistemistica e Comunicazione (DISCo), Universití degli Studi di Milano Bicocca, U14 Viale Sarca 336, I-20126 Milan, Italy;Department of Neurology, Sommelweiss University, 1088 Budapest VIII, Balassa u. 6, Hungary;Department of Radiology, Erasmus MC, University Medical Center, 's-Gravendijkwal 230, 3015 CE Rotterdam, The Netherlands;Department of Information Systems, University of Pannonia, Veszprem, Egyetem u. 10, H-8200, Hungary;Dipartimento di Informatica, Sistemistica e Comunicazione (DISCo), Universití degli Studi di Milano Bicocca, U14 Viale Sarca 336, I-20126 Milan, Italy

  • Venue:
  • Journal of Biomedical Informatics
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

The NEUROWEB project supports cerebrovascular researchers' association studies, intended as the search for statistical correlations between a feature (e.g., a genotype) and a phenotype. In this project the phenotype refers to the patients' pathological state, and thus it is formulated on the basis of the clinical data collected during the diagnostic activity. In order to enhance the statistical robustness of the association inquiries, the project involves four European Union clinical institutions. Each institution provides its proprietary repository, storing patients' data. Although all sites comply with common diagnostic guidelines, they also adopt specific protocols, resulting in partially discrepant repository contents. Therefore, in order to effectively exploit NEUROWEB data for association studies, it is necessary to provide a framework for the phenotype formulation, grounded on the clinical repository content which explicitly addresses the inherent integration problem. To that end, we developed an ontological model for cerebrovascular phenotypes, the NEUROWEB Reference Ontology, composed of three layers. The top-layer (Top Phenotypes) is an expert-based cerebrovascular disease taxonomy. The middle-layer deconstructs the Top Phenotypes into more elementary phenotypes (Low Phenotypes) and general-use medical concepts such as anatomical parts and topological concepts. The bottom-layer (Core Data Set, or CDS) comprises the clinical indicators required for cerebrovascular disorder diagnosis. Low Phenotypes are connected to the bottom-layer (CDS) by specifying what combination of CDS values is required for their existence. Finally, CDS elements are mapped to the local repositories of clinical data. The NEUROWEB system exploits the Reference Ontology to query the different repositories and to retrieve patients characterized by a common phenotype.