The Hilbert marginal spectrum analysis of short-term heart rate signal prior to spontaneous onset of ventricular tachyarrhythmia

The ventricular tachycardia (VT) and ventricular fibrillation (VF) are major causes of triggering sudden cardiac death (SCD) and seriously damage to human life. So, there has a very important significance on prediction research of VT/VF. In this study, the 1024 RR intervals series preceding 135 VT/VF episodes from 78 patients stored by the implantable cardioverter defibrillators (ICD) were analyzed and compared to individually acquired control time series from the same patients normally intrinsic sinus rhythms. The artifacts and ectopic beats were removed by an adaptive filtering algorithm using local variance estimation. After converting RR intervals into instantaneous heart rate (IHR), The Hilbert frequency marginal spectrum of IHR signal were analyzed based on improved Hilbert-Huang transform. Then low-frequency amplitude, high frequency amplitude, very high frequency amplitude, total amplitude and the low-to-high frequency amplitude ratio were extracted from frequency marginal spectrum of IHR signal as signal feature. In comparison with normal sinus rhythm, statistic analysis shows that high frequency amplitude (p<;0.001), very high frequency amplitude (p<;10^-6) and total amplitude (p<;0.002) of heart rate signal before the onset of VT/VF are significantly higher, but the low-to-high frequency amplitude ratio(p<;0.0002)is significantly lower.