Moving Beyond ERP Components: A Selective Review of Approaches to Integrate EEG and Behavior

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Psychiatry

Background:

• Neuroimaging and behavior relationships are key to understanding brain function in health and disease.
• Electroencephalography (EEG) offers a portable, cost-effective measure of brain dynamics but is underrepresented in model-based inference.
• Existing EEG analysis often relies on averaged event-related potentials (ERPs), potentially missing nuanced single-trial information.

Purpose of the Study:

• To bridge the gap between EEG and model-based cognitive inference.
• To highlight advanced EEG analysis techniques for single-trial data.
• To integrate EEG measures with cognitive models for a comprehensive understanding of brain function.

Main Methods:

• Single-trial EEG feature extraction using frequency and time-frequency decomposition.
• Data-driven EEG analysis via blind source separation (BSS) and deep learning.
• Application of cognitive models like drift-diffusion models (DDM) and reinforcement learning (RL).
• Statistical, data-driven, and hierarchical Bayesian models (HBMs) for linking EEG and cognitive variables.

Main Results:

• Demonstration of advanced EEG analysis techniques for enhancing signal quality.
• Integration of cognitive models to extract latent variables from behavioral tasks.
• Establishment of methods to associate EEG-derived features with cognitive model parameters.
• Potential for uncovering moment-to-moment brain dynamics underlying cognition.

Conclusions:

• Advanced EEG analysis and cognitive modeling offer powerful tools for neuroscience research.
• These integrated approaches can provide deeper insights into brain dynamics and cognitive processes.
• The proposed methods are expected to advance theoretical understanding in cognitive neuroscience and inform clinical applications.