Probing neural activations from continuous EEG in a real-world task: time-frequency independent component analysis

Area of Science:

• Neuroscience
• Cognitive Science
• Biomedical Engineering

Background:

• Studying human brain functions during real-world tasks is crucial.
• Neuroimaging techniques like electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) are vital for this research.
• Identifying networked brain activations from continuous EEG data in realistic scenarios presents a significant challenge.

Purpose of the Study:

• To explore the feasibility of using EEG to identify networked brain activations during a realistic task.
• To develop and validate a novel data-driven method for analyzing high-density EEG data.
• To investigate brain processes, their interrelationships, and the effects of sustained attention.

Main Methods:

• A novel time-frequency independent component analysis (tfICA) method was developed.
• tfICA combines time-frequency analysis with complex-valued independent component analysis (ICA).
• High-density EEG data from subjects performing a realistic task were analyzed.

Main Results:

• Six classes of independent components (ICs) with distinct spatio-temporal-spectral patterns were identified across subjects.
• These ICs related to various brain regions including frontal, motor, and occipital cortices.
• Temporal patterns of ICs showed consistency, causal relationships, and correlation with behavioral performance.
• The study observed the time-on-task effect, indicating mental fatigue during a 1-hour sustained task.

Conclusions:

• The tfICA method effectively distinguishes various brain processes from continuous EEG data during realistic tasks.
• Networked brain activations involving visual perception, motor control, working memory, and decision-making were identified.
• The findings demonstrate the potential of tfICA for addressing real-world problems, such as understanding and mitigating time-on-task fatigue.