An ontology-driven architecture for data integration and management in home-based telemonitoring scenarios

  1. Lasierra Beamonte, Nelia
Supervised by:
  1. Alvaro Alesanco Iglesias Director
  2. José García Moros Director

Defence university: Universidad de Zaragoza

Fecha de defensa: 11 December 2012

Committee:
  1. Adolfo Muñoz Carrero Chair
  2. María Canales Compés Secretary
  3. Jorge Enrique López de Vergara Méndez Committee member
  4. Isabel Román Martínez Committee member
  5. Lucy Hederman Committee member

Type: Thesis

Abstract

The shift from traditional medical care to the use of new technology and engineering innovations is nowadays an interesting and growing research area mainly motivated by a growing population with chronic conditions and disabilities. By means of information and communications technologies (ICTs), telemedicine systems offer a good solution for providing medical care at a distance to any person in any place at any time. Although significant contributions have been made in this field in recent decades, telemedicine and in e-health scenarios in general still pose numerous challenges that need to be addressed by researchers in order to take maximum advantage of the benefits that these systems provide and to support their long-term implementation. The goal of this research thesis is to make contributions in the field of home-based telemonitoring scenarios. By periodically collecting patients' clinical data and transferring them to physicians located in remote sites, patient health status supervision and feedback provision is possible. This type of telemedicine system guarantees patient supervision while reducing costs (enabling more autonomous patient care and avoiding hospital over flows). Furthermore, patients' quality of life and empowerment are improved. Specifically, this research investigates how a new architecture based on ontologies can be successfully used to address the main challenges presented in home-based telemonitoring scenarios. The challenges include data integration, personalized care, multi-chronic conditions, clinical and technical management. These are the principal issues presented and discussed in this thesis. The proposed new ontology-based architecture takes into account both practical and conceptual integration issues and the transference of data between the end points of the telemonitoring scenario (i.e, communication and message exchange). The architecture includes two layers: 1) a conceptual layer and 2) a data and communication layer. On the one hand, the conceptual layer based on ontologies is proposed to unify the management procedure and integrate incoming data from all the sources involved in the telemonitoring process. On the other hand, the data and communication layer based on web service technologies is proposed to provide practical back-up to the use of the ontology, to provide a real implementation of the tasks it describes and thus to provide a means of exchanging data. This architecture takes advantage of the combination of ontologies, rules, web services and the autonomic computing paradigm. All are well-known technologies and popular solutions applied in the semantic web domain and network management field. A review of these technologies and related works that have made use of them is presented in this thesis in order to understand how they can be combined successfully to provide a solution for telemonitoring scenarios. The design and development of the ontology used in the conceptual layer led to the study of the autonomic computing paradigm and its combination with ontologies. In addition, the OWL (Ontology Web Language) language was studied and selected to express the required knowledge in the ontology while the SPARQL language was examined for its effective use in defining rules. As an outcome of these research tasks, the HOTMES (Home Ontology for Integrated Management in Telemonitoring Scenarios) ontology, presented in this thesis, was developed. The combination of the HOTMES ontology with SPARQL rules to provide a flexible solution for personalising management tasks and adapting the methodology for different management purposes is also discussed. The use of Web Services (WSs) was investigated to support the exchange of information defined in the conceptual layer of the architecture. A generic ontology based solution was designed to integrate data and management procedures in the data and communication layer of the architecture. This is an innovative REST-inspired architecture that allows information contained in an ontology to be exchanged in a generic manner. This layer structure and its communication method provide the approach with scalability and re-usability features. The application of the HOTMES-based architecture has been studied for clinical purposes following three simple methodological stages described in this thesis. Data and management integration for context-aware and personalized monitoring services for patients with chronic conditions in the telemonitoring scenario are thus addressed. In particular, the extension of the HOTMES ontology defines a patient profile. These profiles in combination with individual rules provide clinical guidelines aiming to monitor and evaluate the evolution of the patient's health status evolution. This research implied a multi-disciplinary collaboration where clinicians had an essential role both in the ontology definition and in the validation of the proposed approach. Patient profiles were defined for 16 types of different diseases. Finally, two solutions were explored and compared in this thesis to address the remote technical management of all devices that comprise the telemonitoring scenario. The first solution was based on the HOTMES ontology-based architecture. The second solution was based on the most popular TCP/IP management architecture, SNMP (Simple Network Management Protocol). As a general conclusion, it has been demonstrated that the combination of ontologies, rules, WSs and the autonomic computing paradigm takes advantage of the main benefits that these technologies can offer in terms of knowledge representation, work flow organization, data transference, personalization of services and self-management capabilities. It has been proven that ontologies can be successfully used to provide clear descriptions of managed data (both clinical and technical) and ways of managing such information. This represents a further step towards the possibility of establishing more effective home-based telemonitoring systems and thus improving the remote care of patients with chronic diseases.