Expert systems for implanted pacemaker analysis.

Abstract

A computer-based software system has been proposed that will assist clinicians in the often difficult task of analyzing multi-channel pacemaker modified ECGs. The system contains three integrated modules. First, digital signal processing indentifies cardiac events from two sources provided by the latest pacemaker technology, intracardiac atrial and ventricular signals, activity markers, and a surface ECG. Secondly, the pacemaker's functional status at each event is analyzed by a rule-based monitor model of the pacemaker. Thirdly, Expert System (ES) techniques were used to combine the digital signal processing metrics, the pacemaker's functional status at each event, and the programmed parameters stored in the pacemaker's memory with a clinical knowledge base relating known problems and their causes. The domain specific tasks, briefly described here, that were most useful in the implementation of the ES are: data reduction, knowledge acquisition methods, knowledge representation, inferencing, and presentation of results.