Dottorato in INGEGNERIA INFORMATICA

Titolo della tesi: Ontology-based Data Management in Healthcare

Nowadays, the study of Big Data has become crucial to guarantee technical advancements in the large majority of the modern computer systems. Within the Healthcare context, this is even a more valid challenge, having the requirement to deploy systems capable of tacking huges amounts of periodically-generated data and with clear standards for assessing data quality, in order to carry on any kind of study. Therefore, the aim of this thesis is to present the results of the application of an Ontology-based data management (OBDM) methodology, developed in the context of the Italian Registry of implantable prostheses (RIPI) project, at National Institute of Health. In particular, its application from scratch to the Italian Arthroplasty Registry (RIAP), one of the registries included in RIPI, is presented. Moreover, the same methodology has been also followed by the Sapienza information-based Technology InnovaTion Center for Health (STITCH) research group to deal with another relevant healthcare domain: the management the Italian Association of Diabetologists (AMD) data. Healthcare domains are usually suitable to be modeled through ontologies, since they follow rules which are well-known by the involved stakeholders. However, clinical data are often provided by several local sources, thus they have to be properly integrated into a central repository in order to be clean and standardized. For this reason, managing this process requires a tight collaboration between computer scientists and healthcare stakeholders, with the purpose of creating systems capable of leveraging the inestimable value of healthcare data through the most suitable approaches of the field. RIAP data layer has been improved up to the point of allowing the project managers to publish a cumulative report on fifteen years of the activity of the registry; AMD dataset was centralized, cleansed and standardized so that several research groups could start to investigate on data, using a data-driven approach to unlock the potential of the involved medical information. From this two real-world experiences, it is possible to conclude that ontologies can enhance healthcare capabilities not only to express the expertise of the clinicians, but also to ease data analytics processes and research.