Some mathematical tools for the analysis of biological signals

The paper presents an informal review of some techniques available for signal analysis. In the interpretation of biomedical signals, the individuation of hidden transient phenomena in the spectrum can have a crucial role for diagnostic purposes. Since most biological signals are nonstationary, the Fourier transform is not sufficient to detect possible transient phenomena in the spectrum; therefore, some improvements in the Fourier transform technique have been carried out by means of window functions in the transformation kernel. Some of the most important features of recent developments in signal analysis are discussed here, with special focus on the uncertainty principle governing any time-frequency analysis.