Spatial-Temporal EEG Imaging for Dual-Loop Neuro-Adaptive Simulation: Cognitive-State Decoding and Communication...

Area of Science:

• Neuroscience
• Human-Computer Interaction
• Cognitive Engineering

Background:

• Human performance degrades in critical environments due to mistimed communication during visual-cognitive saturation.
• Failures result from individual limits and poor team coordination under fluctuating workloads.

Purpose of the Study:

• To present a neuro-adaptive simulation framework using real-time EEG to regulate operator assistance and team communication.
• To decode neurocognitive states and cognitive load for adaptive interventions.

Main Methods:

• Integrated 14-channel wireless EEG, gaze tracking, and communication events via LSL synchronization.
• Employed a hybrid CNN-LSTM model to classify neurocognitive states (Channelized Attention, Diverted Attention, Surprise/Startle) and estimate Cognitive Load Index (CLI).
• Utilized a multi-agent proximal policy optimization (MAPPO) controller for adaptive haptic guidance and a communication gate.

Main Results:

• Achieved over 30% reduction in communication breakdown errors in a dual-station simulation with 25 pilot-engineer pairs.
• Demonstrated strongest effects during peak-load conditions, preserving task progression.
• Improved pilot reaction time and reduced engineer decision load.

Conclusions:

• Spectral-topographic EEG representations provide a practical basis for multimodal neurophysiological sensing and adaptive coordination in human-machine teams.
• Findings show controlled feasibility in simulation, warranting further field validation.
• Emphasized privacy-by-design for operational adaptation, not performance scoring.