Selective windowed time-frequency analysis for the quantitative evaluation of non-stationary cardiovascular signals

A simple and efficient algorithm is developed for the investigation of non-stationary cardiovascular signals. It performs a DFT on windowed parts of the signal, according to the specific frequency and is referred to as Selective Discrete Fourier Transform Algorithm-SDA. It calculates the time-variant spectral components of the signal, displaying a 3-D time-dependent spectrum. The performance of the SDA is tested on simulated data and its capacity to detect the time evolution of the spectral components is proven. Then, the SDA is applied on real HR signal, during vagal maneuvers and transition from supine to upright position. It displays a clear ability to defect and quantitate ANS time-dependent changes. It allows one to analyze and establish the time evolution of the spectral components of the signal, not only during the steady state but also during transient changes