Title :
Knowledge-based approach in the classification of beat waveforms
Author :
Taddei, A. ; Gagliano, R. ; Marchesi, C.
Author_Institution :
CNR Inst. of Clinical Physiol., Pisa, Italy
Abstract :
The authors describe a method for the classification of the beat morphology in long-term electrocardiograms (ECGs). Each QRS-T complex is represented in terms of elementary waveforms extracted from the ECG signal. These waveforms constitute the set of symbols associated with each beat and the related geometrical parameters are their attributes. A knowledge-based system was developed using the Nexpert Object tool for a membership mapping of the ECG beats. Annotated ECG databases (VALE, MIT-BIH) were used for evaluating the classification system
Keywords :
electrocardiography; knowledge based systems; medical diagnostic computing; MIT-BIH; Nexpert Object tool; QRS-T complex; VALE; annotated ECG databases; beat morphology classification; beat waveforms; elementary waveforms; geometrical parameters; knowledge-based system; long-term electrocardiograms; symbols set; Biomedical monitoring; Cardiac disease; Computerized monitoring; Electrocardiography; Morphology; Pathology; Pattern recognition; Physiology; Signal analysis; Spatial databases;
Conference_Titel :
Computers in Cardiology 1991, Proceedings.
Conference_Location :
Venice
Print_ISBN :
0-8186-2485-X
DOI :
10.1109/CIC.1991.168985
