Abstract
Electroencephalography (EEG) records fast-changing neuronal signalling and communication and thus can offer a deep understanding of cognitive processes. However, traditional data analyses which employ the Fast-Fourier Transform (FFT) have been of limited use as they do not allow time- and frequency-resolved tracking of brain activity and detection of directional connectivity. Here, we applied advanced qEEG tools using autoregressive (AR) modelling, alongside traditional approaches, to murine data sets from common research scenarios: (a) the effect of age on resting EEG; (b) drug actions on non-rapid eye movement (NREM) sleep EEG (pharmaco-EEG); and (c) dynamic EEG profiles during correct vs incorrect spontaneous alternation responses in the Y-maze. AR analyses of short data strips reliably detected age- and drug-induced spectral EEG changes, while renormalized partial directed coherence (rPDC) reported direction- and time-resolved connectivity dynamics in mice. Our approach allows for the first time inference of behaviour- and stage-dependent data in a time- and frequency-resolved manner, and offers insights into brain networks that underlie working memory processing beyond what can be achieved with traditional methods.
Highlights
Tracking transient patterns of connectivity in defined behavioural or disease states is one of the most pressing challenges in neuroscience
We demonstrate that AR and Fast Fourier transform (FFT) are equivalent for a chosen experimental example and confirm that the additional directional information provided by renormalized partial directed coherence (rPDC) is congruous with results obtained via classical coherence
These approaches were limited in their temporal resolution and only applied to bivariate analyses
Summary
Tracking transient patterns of connectivity in defined behavioural or disease states is one of the most pressing challenges in neuroscience. Functional connectivity analysis of EEGs is usually based on correlations (coherence) between signals. Spectral analysis of time series is commonly based upon Fourier transformation, typically employing the Fast Fourier transform (FFT). To achieve consistent spectral estimates, windowing of the time series or smoothing of the FFT is required. By embedding the AR model into a state space model and augmenting it by adding an equation for the changes in parameters of the AR process, it is possible to estimate parameters that change over time as long as they change more slowly than the (stochastic) dynamic itself. The state space model allows one to account for observational noise. Using the Expectation-Maximization algorithm applying the dual Kalman filter, parameters can be estimated in a time-resolved manner in the presence of observational noise[9]. A multivariate generalization of the AR model enables the analysis of interdependency between EEG signals
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call