| Abstract |
The evolution of the World Wide Web has transformed Internet into a context capable for information sharing among its users. The information in the various resources over the web is characterized by its heterogeneity and its distributed nature. The problem of finding ways and methodologies for interconnection over distributed and heterogeneous sources of information and data have become critical, particularly in fields where the volume of the collected information and data is rapidly increasing. An area that has experienced radical change, and continues to do so, is health-care delivery. The requirements in this field increase the need for integration over the distributed and heterogeneous clinical data sources in a way that: the direct access in data, and the reliable support of clinical decision-making processes (evidence-based clinical decision making) may help experts to orient themselves in the information space and enhance their decision-making capabilities. The access at information and the collection of data are not end in itself. What is desirable is the exploitation of data, hence the possibility for exporting useful and comprehensible conclusions. Focusing in the field of health, and within the framework of HygeiaNet -The Integrated Health Care Network of Crete (www.hygeianet.gr), we try to give impulse to a new direction of clinical data management, as retrieved from the Integrated Electronic Health Care Record (IEHCR). From patient-oriented view we pass into a population-oriented view. This shift will offer the potential of Internet-based epidemiology and the exploitation of the relative results in the daily clinical practice. The main challenge is the operational incorporation of data mining mechanisms for the shake of knowledge discovery from distributed and heterogeneous information sources. Towards this end, a multi-layer data integration and processing approach is proposed that deals with: a) the smooth and seamless semantic homogenisation of heterogeneous data, b) the development and adaptation of knowledge discovery operations, c) the presentation of results via visualization techniques and tools, and d) development of operational and user friendly interface (human computer interface - HCI). The final product is a flexible and functional environment for the standardisation and the intelligent mining of clinical data, the HealthObs system.
|
Ένα Ολοκληρωμένο Σύστημα Εξόρυξης Δεδομένων και Ανακάλυψης Γνώσης από Κατανεμημένες και Ετερογενείς Κλινικές Πηγές Δεδομένων
