Interpretable EEG cognitive state assessment using fuzzy modeling and case-based neurophysiological evidence

Area of Science:

• Neuroscience
• Cognitive Science
• Biomedical Engineering

Background:

• Real-time cognitive-state assessment is crucial for adaptive learning, neuroergonomics, and safety-critical systems.
• Electroencephalography (EEG) offers a method for monitoring neural oscillations, reflecting changes in attention and working memory.
• Existing methods may lack interpretability or real-time dynamic assessment capabilities.

Purpose of the Study:

• To propose an interpretable EEG-based framework for real-time cognitive-state assessment.
• To dynamically profile attentional functions and working memory volumes using sensitive EEG measures.
• To develop a transparent fuzzy-logic mechanism for representing cognitive status perception.

Main Methods:

• Utilized EEG signals from bipolar pairs F7-T3 and F8-T4 for verbal and visuospatial working memory.
• Employed T4-T6 EEG monitoring for attentional engagement across different activity types.
• Developed a dynamic weighting model to generate a composite working-memory index and integrated it with attention measures for a continuous cognitive state index (CSI).

Main Results:

• The framework successfully generated a continuous cognitive state index (CSI) indicating passive, moderate, and focused engagement.
• Case studies demonstrated clear neurophysiological patterns correlating with cognitive load and memory maintenance.
• Identified specific EEG patterns: right fronto-temporal fluctuations for high cognitive load, stable temporo-parietal activity for memory maintenance, and left frontal steady activity.

Conclusions:

• The proposed EEG framework provides an interpretable and effective method for real-time cognitive-state assessment.
• The system's ability to track dynamic changes in working memory and attention is validated through case studies.
• This approach holds potential for enhancing adaptive learning systems, neuroergonomics, and monitoring cognitive status in critical applications.