A novel pattern mining approach for identifying cognitive activity in EEG based functional brain networks.

Area of Science:

• Neuroscience
• Computational Biology
• Data Mining

Background:

• Understanding complex neuronal interactions in the human brain is challenging.
• Innovative mathematical and computational models are needed to study brain region activity during cognitive tasks.

Purpose of the Study:

• To present a novel Common Functional Pattern Mining approach.
• To demonstrate similar interaction patterns arising from common behaviors of brain regions.

Main Methods:

• Modeled electroencephalography (EEG)-based functional brain network electrode sites as transactions.
• Represented node-based complex network measures as itemsets.
• Transformed itemsets into a Functional Pattern Graph for mining.

Main Results:

• Identified common functional patterns related to specific brain functioning.
• Empirically demonstrated the approach's efficiency in detecting similarities between electrode sites.
• Showcased pattern identification across various cognitive load states.

Conclusions:

• The Common Functional Pattern Mining approach effectively reveals underlying mechanisms of neuronal activity.
• This method aids in understanding brain region interactions during different cognitive states.
• The approach provides insights into functional brain network dynamics.