Task-Independent Mental Workload Classification Based Upon Common Multiband EEG Cortical Connectivity.

Area of Science:

• Neuroscience
• Cognitive Science
• Brain-Computer Interfaces

Background:

• Classifying mental workload is crucial in neuroscience but challenging across different tasks.
• Network approaches offer promising insights into brain organization and mental states.

Purpose of the Study:

• To develop a framework for discriminating mental workload within and across tasks using brain connectivity.
• To identify key neural mechanisms underlying different cognitive tasks.

Main Methods:

• Utilized multiband electroencephalography (EEG) to construct functional brain networks in source space across various frequency bands.
• Employed a sequential feature selection algorithm to identify salient functional connectivity features.
• Applied these features for classification of mental workload in N-back and mental arithmetic tasks.

Main Results:

• Achieved high classification accuracy: 87% for cross-task, 88% for N-back, and 86% for mental arithmetic.
• Identified a small subset of discriminative functional connections.
• Found that frontal areas in theta and beta bands were key for classification.

Conclusions:

• The proposed method effectively classifies mental workload within and across tasks using EEG functional connectivity.
• The identified connectivity patterns highlight shared and distinct neural mechanisms between cognitive tasks.
• This approach advances the understanding of brain function during cognitive load.