[Brief causal relations in EEG based on adaptive Granger causality]

Area of Science:

• Neuroscience
• Signal Processing
• Cognitive Science

Background:

• Traditional EEG analysis methods like coherence and correlation cannot determine the direction of signal interrelations.
• Establishing causality is crucial for understanding brain network dynamics.
• Granger causality offers a framework for inferring causality from time-series data.

Purpose of the Study:

• To develop and apply an adaptive estimation of Granger causality for detecting dynamic causal relations in EEG signals.
• To investigate directed neural interactions during cognitive tasks, specifically the Stroop task.
• To analyze causal relationships within short time intervals (less than 100 ms).

Main Methods:

• Adaptive estimation of Granger causality.
• Time-variant Granger causality analysis.
• Application to electroencephalography (EEG) data during the Stroop task.

Main Results:

• The adaptive Granger causality method successfully detected dynamic causal relations within short time windows (<100 ms).
• Analysis of Stroop task EEG data revealed a dense web of directed interactions.
• These interactions predominantly flowed from posterior to anterior cortical areas during conflict situations.

Conclusions:

• Adaptive Granger causality is a powerful tool for uncovering time-varying causal relationships in neural data.
• Cognitive conflict during the Stroop task induces significant directed information flow within the brain.
• The findings highlight a posterior-to-anterior cortical communication pattern associated with conflict processing.