Designing an alarm system for the stratification of risk of cardiac arrhythmias

The idea of this research is to propose an alarm system that can alert the early establishment of an event paroxysmal ventricular tarquicardia (PVT), from study of heart rate variability (HRV). The study of the variability is support in the frequency of abnormal patterns prior to the events. We select and analyze in detail 24 hours records from dynamic electrocardiography (Holter). We selected 88 one-hour segments prior to the events of PVT, and once analyzed the system found an efficiency of 77% in early detection of the establishment of the arrhythmia, with an average prediction time of 43±8 minutes distant from events, statistically significant result. To assess the percentage of false alarms we evaluate 103 hours from control healthy subjects, resulting in a specificity of 85%.