On the evaluation of heart rate spectra: the Lomb periodogram

The Lomb periodogram is a method for the spectral analysis of randomly sampled data. For this property and for the ability to estimate spectral peaks beyond the mean Nyquist frequency, the Lomb approach was proposed for the spectral analysis of heart rate (HR). The features of the Lomb periodogram, however hold for a random sampling while the sampling frequencies of the HR series are equal to the same HR values and thus include systematic oscillations and 1/f noise typical of HR variability. The aim of this study is to investigate whether the rhythmic oscillations and the 1/f noise may influence the performances of the Lomb method. This is done by applying the Lomb procedure to simulated HR series and to real HR data. Results indicate that the valuable features of the Lomb periodogram are in part nullified by the specific characteristics of the HR signal.