Stefania Montani, Riccardo Bellazzi, Alberto Riva, Cristiana Larizza, Luigi Portinale, Mario Stefanelli
We present a successful application of Artificial Intelligence (AI) methodologies in the context of a telemedicine service for diabetic patients management, developed within the EU-funded T-IDDM project. The system architecture is distributed, and composed by a Patient Unit and by a Medical Unit, connected through a telecommunication link. Several AI methods have been exploited to implement the T-IDDM functionality. The data base relies on an explicit representation of the domain ontology. Temporal Abstractions and other Intelligent Data Analysis techniques are used to analyse the patient's monitoring data; the Case Based Reasoning (CBR) methodology is applied to perform the Knowledge Management task. Finally, CBR is integrated with Rule Based Reasoning to provide physicians with a multi-modal reasoning decision support tool. The T-IDDM service is being tested through a small on field trial in Pavia; the first results, though preliminary, seem to substantiate the hypothesis that the use of an AI-based telemedicine system could present an advantage in the management of type 1 diabetic patients, leading to a more tight control of the patients' metabolic situation, in a cost-effective way.
Keywords: Intelligent Data Analysis, Case Based Reasoning, Knowledge Based Systems, Telemedicine, Diabetes
Citation: Stefania Montani, Riccardo Bellazzi, Alberto Riva, Cristiana Larizza, Luigi Portinale, Mario Stefanelli: Artificial Intelligence Techniques for Diabetes Management: the T-IDDM Project. In W.Horn (ed.): ECAI2000, Proceedings of the 14th European Conference on Artificial Intelligence, IOS Press, Amsterdam, 2000, pp.716-720.