Multiagent systems for cardiac pacing simulation and control

  • Authors:

  • Francesco Amigoni;Alessandro Beda;Nicola Gatti

  • Affiliations:

  • Artif. Intell. and Robotics Lab., Dipartimento di Elettronica e Informazione, Politecnico di Milano, 20133 Milano, Italy, amigoni@elet.polimi.it (Corresp. Dipartimento di Elettronica e Informazion ...;Institute of Sound and Vibration Research, Southampton University, University Road, Highfield, Southampton S017, 3RZJ, UK E-mail: ab1@isvr.soton.ac.uk;Artificial Intelligence and Robotics Laboratory, Dipartimento di Elettronica e Informazione, Politecnico di Milano, Piazza Leonardo da Vinci 32, 20133 Milano, Italy, ngatti@elet.polimi.it

  • Venue:
  • AI Communications - Agents Applied in Health Care
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

Simulating and controlling physiological phenomena are complex tasks to tackle. This is due to the fact that physiological processes are usually described by a set of partial models representing specific aspects of the phenomena and their adoption does not allow the achievement of an effective simulation/control system. A current open issue is the development of techniques able to comprehensively describe a phenomenon exploiting partial models. Simulation and control heavily rely on accurate modelling of physiological systems. In addition, since a large number of partial models of a single physiological phenomenon have been proposed over the years, the evaluation of their effectiveness and of their combinations is a fundamental task. In this paper we propose a multiagent paradigm, called anthropic agency, as a flexible tool to support and evaluate the combination of partial models embedded in agents. We present an agent negotiation paradigm, that improves the one we employed in our previous applications, as a flexible approach to combine optimally the partial models. We formally describe the negotiation protocol and we embed it in a FIPA agent interaction protocol. Furthermore, as an example of practical application, we describe how our paradigm can be a potential solution to the problem of adaptive cardiac pacing. Finally, we experimentally evaluate our approach and we discuss its properties and peculiarities.