ABSTRACT
Currently, the treatment and prevention of cardiac arrhythmias include antiarrhythmic drug therapy, surgery, ablation, and electronic intervention. Antiarrhythmic drugs, which traditionally have been the only method of therapy for many life-threatening arrhythmias, are often unreliable and unsafe in the prevention of tachyarrhythmias. Surgery, such as excision or cryoblation, is considered appropriate only for a small number of patients and is highly invasive. Ablation has become an excellent means for eliminating a number of arrhythmias with well-defined electrophysiological mechanism. However, there remain hundreds of thousands of individuals for whom the mechanism of life-threatening arrhythmia is ill-defined and thus difficult to treat using traditional measures. For these individuals, implantable electronic antiarrhythmic devices, such as antitachycardia pacemakers, cardioverters, and automatic defibrillators, have become the therapy of choice. Other implantable devices, such as drug-infusion pumps, have also been developed [1] and show promise for future antiarrhythmic therapy. Such devices offer a variety of therapies, each tailored to a specific arrhythmia. Implantable devices are expected not only to administer appropriate therapy in the presence of an arrhythmia, but also to monitor the heart for extensive periods and recognize the need to administer preventive therapy before severe
dysfunction occurs [2,3]. Detecting the presence of an arrhythmia requires these devices to have automated detection schemes with high sensitivity and specificity that are immune to changing clinical conditions, such as changes in drug therapy, lead configuration, and physical activity. Automated arrhythmia recognition requires a quantitative description of arrhythmias in terms of standard signal analysis and processing techniques.
