Smoothed pseudo Wigner–Ville distribution as an alternative to Fourier transform in rats

Abstract

Techniques for examining signals in the time and frequency domains are well-established tools. These tools have their limitations; they tell us in a broad sense where the signal component exists in the frequency domain, but they do not tell us how its frequency characteristics change over time. The time-frequency has become a powerful alternative for the analysis of signals. Among various time-frequency distribution methods, one of the most studied is the Wigner–Ville distribution. The aim of this study was to evaluate in conscious rats smoothed pseudo Wigner–Ville distribution (SPWVD) as an alternative to the fast Fourier transform (FFT) in RR intervals and in systolic blood pressure (SBP), before and after adrenergic and cholinergic receptor blockade. Fourteen Wistar rats equipped with telemetry probe were evaluated: (1) under control conditions; (2) after injection of saline (100 μl kg −1 i.v.); (3) after atenolol (1 mg kg −1 i.v.); (4) after atropine methyl nitrate (0.5 mg kg −1 i.v.); and (5) after phentolamine (5 mg kg −1 i.v.). FFT and SPWVD were applied to RR intervals and SBP time series. Six-minute time series of RR intervals, systolic and diastolic pressures were analysed. The bias and distribution of differences between FFT and SPWVD methods in RR intervals under base conditions were 1.4±0.4% ( r 2=0.94; P<0.01) in LF/LF+HF; 1.5±0.5% ( r 2=0.92; P<0.01) in HF/LF+HF and 4.8±1.9% ( r 2=0.92; P<0.01) in LF/HF. In SBP the bias and distribution were 1.5±0.8% ( r 2=0.90; P<0.05) in LF/LF+HF and 1.7±0.6% ( r 2=0.92; P<0.01) in HF/LF+HF. In the frequency domain analysis of RR intervals and SBP there was no difference between FFT and SPWVD. The agreement between the methods demonstrates that in stationary signals both methods can be used interchangeably. SPWVD may be an interesting tool to analyse biomedical signals; it provides a good resolution at high frequency and a good frequency resolution at low frequencies independently if signals remain stationary.