A Subproject of the German Priority Research Program on "Intelligent Agents and Realistic Commercial Application Scenarios"

MedPAge (Medical Path Agents)

Situation

The increasing cost pressure combined with the constraint for economisation imposes big problems on the organisation of hospital processes, because diagnosis and treatments have to be carried out under consideration of spontanous or at least fast response times and anticipatory planning. It is an enormous challenge for the planning systems to achieve an optimal solution with respect to the possibly emerging goal conflicts. The present systems suffer from the sparse specified decision paths and flexible processes that can occur during the treatments.

Approach

The MedPAge project propses a new solution attempt by combining two approaches: The concept of standardized clinical pathways on the one hand and the flexible coordination mechanisms of multi-agent-systems on the other hand embodied in a consolidated dynamic system architecture shall lead to improved and more efficient planning results and help mastering the problem domain. The verification of the predominance of this combination shall be shown by simulation methods close to reality and practical tests. These will further reveal possible problems and weaknesses and lead to a better understanding of hospital processes. Furthermore the system can be used to evaluate the usage of other agent based agent approaches in health care. Unique selling point of the MedPAge system is the decentralized and patient centered point of view, which shall contribute especially to a more patient friendly scheduling.

In the course of the DFG founded MedPAge research project a multi agent system is designed with the objective to utilize decentralized agent-oriented methods for planning and coordination of hospital activities, and to verify the practical relevance within simulation experiments. Especially, the conceptual development and the experiments shall show, how the use of autonomous, negotiating agents can improve the management of hospital processes. The presented objective is realized in an interdisciplinary cooperation of the Distributed Systems and Information Systems Group of the University of Hamburg and of the Business Management and Business Informatics Group of the University of Mannheim (formerly Business Informatics Group of the University of Bayreuth).

funded by the DFG