Using scalar fuzzy control for aggregation problems in knowledge based diagnosis for implantable devices

The Scalar Fuzzy Control (SFC) is a newly suggested extension to the Fuzzy Set Theory, providing the possibility to easily represent and implement unprecise human knowledge. In this paper we discuss how the SFC can be used in the area of aggregation problems, combining the SFC with conventional aggregation techniques. We first give a short summary on the SFC, then outline the area of aggregation problems and afterwards discuss the technique to integrate both methods. In the second part of this paper we discuss these theoretical results on the basis of an expert system used to detect and deal with changing operating conditions of Implantable Cardioverter Defibrillators (ICDs). Changes include lead dislodgment, increased pacing thresholds, inadequate therapy due to changed medical conditions with the patient, etc. It is the task of the expert system to detect such changes, to identify the root cause and to propose mitigation. We show that the introduction of the SFC into aggregation problems is an effective method to reduce the efforts for knowledge base implementation and maintenance.