Modeling Brain Waves as a Mixture of Latent Processes

The standard approach to analyzing brain electrical activity is to examine the spectral density function (SDF) and identify predefined frequency bands that have the most substantial relative contributions to the overall variance of the signal. However, a limitation of this approach is that the precise frequency localization and bandwidth of oscillations vary with cognitive demands, thus ideally should not be defined \emph{a priori} in an experiment. In this paper, we develop a data-driven approach to identifies (i) the number of prominent peaks, (ii) the frequency peak locations, and (iii) their corresponding bandwidths (or spread of power around the peaks). We propose a Bayesian mixture auto-regressive decomposition method (BMARD), which represents the standardized SDF as a Dirichlet process mixture based on a kernel derived from second-order auto-regressive processes characterized by location (peak) and scale (bandwidth) parameters. We present a Metropolis-Hastings within Gibbs algorithm to sample from the posterior distribution of the mixture parameters. Simulation studies demonstrate the robustness and performance of the BMARD method. Finally, we use the proposed BMARD method to analyze local field potential (LFP) activity from the hippocampus of laboratory rats across different conditions in a non-spatial sequence memory experiment to examine the link between specific patterns of activity and trial-specific cognitive demands.