Detection of supraventricular and ventricular ectopic beats using a singled lead ECG

Two simple algorithms for supraventricular (SVEB) and ventricular ectopic beat (VEB) detection using the electrocardiogram (ECG) are presented. Both algorithms use time-domain features and a linear classifier. The first algorithm requires QRS detection only and the second algorithm requires P, QRS and T wave segmentation. Data was obtained from the 44 non-pacemaker recordings of the MIT-BIH arrhythmia database and contained approximately 100,000 beats. Performance assessment of the best system resulted in an accuracy of 94.4% when discriminating SVEB from non-SVEBs and 97.8% in discriminating VEB from non-VEBs