Time-Frequency Analysis of Analog Neural Signals

Definition

Time-frequency analysis is a signal processing method that involves extracting frequency-specific information from temporally localized windows of a signal. Time-frequency analysis is most useful for interpreting signals that are nonstationary, meaning that the spectral characteristics change over time.

Detailed Description

Why Spectral Analysis?

Population-level neural activity, as reflected by the local field potential and electroencephalogram, often exhibits rhythmic temporal patterns with characteristic frequencies, for example, at 10 Hz (“alpha”) or 40 Hz (“gamma”). These rhythmic patterns are called neural oscillations and have been linked to myriad perceptual, cognitive, and motor phenomena (Buzsáki 2006). Careful inspection of rhythmic neural time series suggests that these rhythms are not purely sinusoidal (Cole and Voytek 2017; Jones 2016) yet are sinusoidal enough to justify analysis methods that use sine waves as basis functions, such as the Fourier transform...