Algorithms for fibrillation and tachyarrhythmia detection.

Sudden cardiac death (SCD) is the most common mode of death in the United States, affecting approximately 450,000 people per year. One therapy that has proved effective in managing ventricular arrhythmia recurrences (ventricular tachycardia-VT; ventricular fibrillation-VF) in patients surviving SCD is the AICD Automatic Implantable Cardioverter Defibrillator system. Winkle reported a 98.7% sudden cardiac death survival with the AICD device at 1 year and 94% at 5 years,’ compared to the expected survival of 70% at 1 year for untreated SCD survivors.2 Much work has been focused recently on AICD algorithms for the electrical treatment of ventricular arrhythmias precipitating sudden cardiac death. Arrhythmia detection for the current AICD device is based on a single ventricular rate threshold (with ECG morphology as an optional feature). Although this straightforward approach to therapy guidance yields a high sensitivity to VTNF, the conservative stance of the rate threshold criterion could cause undesired ventricular therapy for nonventricular rhythms (sinus or supraventricular tachycardia) . Current morphologic detection enhancements can reduce this occurrence in patients with narrow sinus QRS complexes. The challenge in designing future algorithms for automatic treatment of ventricular arrhythmias is to improve the specificity of the implanted device to lethal arrhythmias without sacrificing its sensitivity. The device must be able to identify the type of tachycardia and to deliver therapy when needed. According to the literature, therapy control algorithms have been based on three approaches: analysis of rate, timing, or morphology.