Preprocessing effects in time–frequency distributions and spectral analysis of heart rate variability

Heart rate variability (HRV) is very significance noninvasive tool for autonomic nervous system (ANS) analysis. HRV signal includes both slowly changing components and rapidly changing transient events. This study presents effects of preprocessing of HRV in time–frequency analysis and spectral estimations. Preprocessing includes two levels as detrending of trend using smoothness prior method and correction of ectopics using integral pulse frequency modulation (IPFM). The datasets used in this study are obtained from the Spontaneous Ventricular Tachyarrhythmia (VTA) database. Datasets include least one ventricular tachyarrhythmia (VT) or ventricular fibrillation (VF) episode. Effects of preprocessing are investigated for time–frequency analysis using continuous wavelet transform (CWT) and spectrogram and for spectral analysis using periodogram, Welch's periodogram and Burg's periodogram. Performance of these methods in determination of VT or VF episode is analyzed. Importance of preprocessing is explained comparing of obtained results.