High-frequency Broadband Modulations of Electroencephalographic Spectra

Area of Science:

• Neuroscience
• Cognitive Science
• Signal Processing

Background:

• High-frequency cortical potentials in electroencephalography (EEG) are difficult to isolate due to low amplitudes and confounding scalp muscle activity.
• Existing methods struggle to differentiate neural signals from artifacts in high-frequency EEG bands (beta, gamma, high gamma).

Purpose of the Study:

• To develop and validate a method for isolating distinct high-frequency electroencephalographic (EEG) signal sources.
• To investigate the relationship between isolated neural activity modulations and cognitive processes, specifically emotion imagination.

Main Methods:

• Applied independent component analysis (ICA) to decompose electroencephalographic (EEG) data from an emotion imagination task.
• Utilized joint decomposition of independent component (IC) log spectrograms to identify source- and frequency-independent modulators (IMs).
• Employed multi-dimensional scaling to analyze relationships between brain IM effects and emotional valence.

Main Results:

• Identified three distinct classes of independent modulators (IMs) that separately control high-frequency power (15-200 Hz) in brain, scalp muscle, and ocular motor IC processes.
• Demonstrated that these IMs can be isolated from scalp-recorded EEG, contrary to prevalent assumptions.
• Revealed significant, spatially complex associations between broadband brain IM effects and the valence of imagined emotions.

Conclusions:

• Unitary modes of spectral modulation within beta, gamma, and high gamma frequencies can be effectively isolated from scalp EEG.
• These isolated modulations are differentially associated with distinct neural sources and cognitive activities like emotion processing.
• The findings challenge the assumption that high-frequency EEG signals are inherently inseparable and offer new avenues for cognitive neuroscience research.