Author/Editor		Jamšek, Janez; Stefanovska, Aneta
Title		Selektivna diskretna Fourierova analiza
Translated title		Selective discrete Fourier analysis
Type		članek
Source		In: Zajc B, editor. Zbornik 7. elektrotehniške in računalniške konference ERK'98. Zvezek B. Računalništvo in informatika, umetna inteligenca, robotika, razpoznavanje vzorcev, biomedicinska tehnika, močnostna elektrotehnika, didaktika, študentski članki; 1998 sep 24-26; Portorož. Ljubljana: Slovenska sekcija IEEE,
Publication year		1998
Volume		str. 339-42
Language		slo
Abstract		The cardiovascular related signals contain several basic frequencies with a ratio of 1:100 between the highest and the lowest frequency (1,2). Moreover, the values of the basic frequencies and their corresponding amplitudes vary in time. When a signal contains several different basic frequencies that are variable in time, the spectral analysis using Fourier transform will smear the time variations of the spectral components over the entire duration of the trace. Recently, an algorithm that enables selective frequency resolution for each frequency component was introduced (3), named selective discrete Fourier transform algorithm (SDFTA). It is an extension of the short time Fourier transform (STFT) . It is extension of the short time Fourier transform (STFT) (4) and is based on discrete Fourier transform (DFT). We present the characteristics of the SDFTA. Its potentials are demonstrated on a set of numerically generated signals, similar to measured cardiovascular signals. Time-frequency representations, obtained by SDFTA and STFT, are compared. Using the SDFTA each spectral component is calculated from a different length of signal, while in STFT a constant length of signal is used. It is shown that the SDFTA provides good time resolution for high frequencies and also good frequency resolution for low frequencies. The SDFTA was also used to estimate the time-frequency distribution of a signal of respiratory activity measured on a healthy, resting subject. It is shown that the changes in time of its basic frequency, of around 0.3 Hz, and the corresponding amplitude can be well represented.
Descriptors		FOURIER ANALYSIS ALGORITHMS SIGNAL PROCESSING, COMPUTER-ASSISTED RESPIRATION SOUND SPECTROGRAPHY