Towards a Knowledge Acquisition and Management System for ECG Diagnosis
Ph. Bourlas, Sgouros N.M., GEORGE PAPAKONSTANTINOU, Panayiotis D. Tsanakas · Studies in health technology and informatics · 1996
We propose a knowledge acquisition and management system for 12-lead electrocardiogram (ECG) diagnosis. This system uses an extended form of Quinlan's ID3 algorithm to learn rules, synthesizing decision trees by inductive inference from examples. Furthermore, the system includes a knowledge management component that continuously optimizes the knowledge base by learning new rules and merging them with the old ones. We have developed a set of ECG attributes on which the classification of the training set and the extraction of the diagnosis rules are based. An expert system for ECG diagnosis provides the initial knowledge base for this diagnosis environment.