Estimating Directed Connectivity from Cortical Recordings and Reconstructed Sources

Area of Science:

• Cognitive Neuroscience
• Neurophysiology
• Biophysics

Background:

• Electrical brain activity is commonly recorded non-invasively via scalp electroencephalography (EEG).
• Reconstructing source-level brain activity from EEG is crucial for understanding neuronal contributions and connectivity.
• A validated method for assessing the accuracy of EEG-based source reconstruction is needed.

Purpose of the Study:

• To validate source-level connectivity reconstructed from EEG against invasive electrocorticography (ECoG) recordings.
• To assess how frontal-parietal brain connectivity changes between wakefulness and sedation.
• To compare connectivity results from invasive ECoG and reconstructed EEG sources.

Main Methods:

• Simultaneous non-invasive EEG and invasive ECoG recordings were obtained from a macaque monkey.
• Directed connectivity in the frequency domain was analyzed using data-driven (directed transfer function) and model-driven (dynamic causal modelling) approaches.
• Connectivity was assessed between frontal and parietal cortical sources under wakefulness and sedation.

Main Results:

• Directed connectivity between cortical sources reconstructed from scalp EEG was qualitatively similar to that inferred from direct cortical recordings.
• Connectivity analyses using both invasive ECoG and reconstructed EEG sources yielded consistent results.
• Changes in connectivity between frontal and parietal sources were consistent with previous findings and modulated by sedation.

Conclusions:

• Inferences about directed brain connectivity derived from non-invasive EEG data demonstrate construct validity when compared to invasive ECoG recordings.
• The study provides a valuable dataset for validating source reconstruction techniques in cognitive neuroscience.
• EEG-based connectivity analysis is a reliable method for studying brain network dynamics.