EVOLUTIONARY MAXIMUM-ENTROPY SPECTRAL ESTIMATION AND HEART-RATE-VARIABILITY ANALYSIS

Spectral analysis has been used extensively in heart rate variability (HRV) studies. The spectral content of HRV signals is useful in assess ing the status of the autonomic nervous system. Although most of the H RV studies assume stationarity, the statistics of HRV signals change w ith time due to transients caused by physiological phenomena. Therefor e, the use of time-frequency analysis to estimate the time-dependent s pectrum of these non-stationary signals is of great importance. Recent ly, the spectrogram, the Wigner distribution, and the evolutionary per iodogram have been used to analyze HRV signals. In this paper, we prop ose the application of the evolutionary maximum entropy (EME) spectral analysis to HRV signals. The EME spectral analysis is based on the ma ximum entropy method for stationary processes and the evolutionary spe ctral theory. It consists in finding an EME spectrum that matches the Fourier coefficients of the evolutionary spectrum. The spectral parame ters are efficiently calculated by means of the Levinson algorithm. Th e EME spectral estimator provides very good time-frequency resolution, sidelobe reduction and parametric modeling of the evolutionary spectr um. With the help of real HRV signals we show the superior performance of the EME over the earlier methods.