Abstract
dc:descriptionMany cardiac diseases are treated with electronic implants today. Typical implants are pacemakers and implantable cardioverter defibrillators (ICD). As a preferable therapeutic treatment the implants have been developed to highly programmable electronic devices in the recent decades. CWA GmbH, Aachen, Germany, develops in cooperation with the „Innere Medizin I” of the clinical center in Aachen, Germany, and the BIOTRONIK GmbH, Berlin, Germany, expert systems for programming such devices, enabling the physicians to program the devices by using their medical jargon rather than in terms of technical parameters. Thus, devices programming without detailed technical know how on the devices is possible. Topic of this thesis is the modeling and implementation of unprecise human expert knowledge within expert systems. Existing techniques are presented, above all the Bayes theorem, the Dempster Shafer theory and the theory of Certainty Factors. As one of the most important techniques to implement unprecise knowledge the Fuzzy Set Theory and Fuzzy Control in particular are discussed more detailed. The Fuzzy methods do not only distinguish themselves as important and suitable techniques, but show up with some problems when dealing with huge knowledge bases with hundreds of rules too. These problems are presented and discussed in detail. A new idea of this thesis is an extension to the existing Fuzzy Set Theory, which provides the possibility to efficiently implement, verify and validate unprecise knowledge in huge knowledge bases, while simultaneously increasing the accuracy (in the sense of the formulated human knowledge) and reducing the required calculation power. The new technique is called Scalar Fuzzy Control. It provides the possibility to directly implement human expert knowledge which deals with real data values (i.e. scalar values), instead of transforming them into multi-dimensional linguistic variables and matrices. In order to implement unprecise knowledge on the scalar values, unprecise comparison and combination operators are derived and defined. Their properties are discussed in detail and important theorems are set up and proofed. The subject of expert systems for programming pacemakers and ICDs is used to demonstrate the practical usage of the new technique. The thesis gives a short introduction to cardiac diseases and therapeutic treatments first. By using the Scalar Fuzzy Control the implementation of unprecise knowledge is demonstrated and the results are compared to implementations using the „conventional” Fuzzy Control. The implementation was inspected and proved in cooperation with physicians from the „Innere Medizin I” of the clinical center in Aachen. All in all it is shown that the Scalar Fuzzy Control is a valuable technique to implementing unprecise knowledge.
Degree
thesis:*- Grantor dc:publisher
- Klinkenberg
- Year dc:date
- 2003
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Mlynski, Michael Frank
- Contributors dc:contributor
-
- Ameling, Walter
Subjects
dc:subject × 6Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- ger
Identifiers
dc:identifier.*- OAI identifier oai:identifier
- oai:publications.rwth-aachen.de:61526