Ventricular arrhythmia detection using time-domain template algorithms

Implantable cardioverter defibrillators (ICDs) have shown success in reliably terminating potentially fatal arrhythmias, i.e. ventricular tachycardia (VT) and ventricular fibrillation (VF), by delivering electrical therapy to restore normal sinus rhythm. However, the ICD often applies defibrillation to VT instead of employing specialized lower energy therapies. In this study, variability (standard deviation and interquartile range) of four proposed algorithms was tested to determine their capacity to discriminate VT from VF. Algorithms included correlation waveform analysis, bin area method, normalized difference of area method, and derivative area method. Multiple thresholds were tested and relative operating characteristic curves compared parameters and algorithms. Results indicated that these methods were able to distinguish between VT and VF in most patients. However, due to loss of VF sensitivity, the algorithms in their current form are not recommended for implementation in commercial devices