Cognitive Load Estimation Using Brain Foundation Models and Interpretability for BCIs

2026-01-29

AI Summary

This work addresses the limitations of traditional EEG approaches in handling inter-subject variability and task-specific preprocessing by introducing brain foundation models (BFMs) for the first time to long-term cognitive workload monitoring. By fine-tuning only a few network layers, the method extracts generalizable features to enable high-accuracy, real-time cognitive load estimation. Furthermore, interpretability analysis via Partition SHAP reveals the pivotal role of the prefrontal cortex in cognitive control and captures longitudinal neural dynamics throughout the learning process. The proposed approach outperforms current state-of-the-art methods, offering superior performance while maintaining real-time capability, cross-subject generalizability, and alignment with established neurocognitive mechanisms.

Technology Category

• Cognitive Modeling & Cognitive Systems: Adaptive Behavior
• Humans and AI: Brain-Sensing and Analysis
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Accurately monitoring cognitive load in real time is critical for Brain-Computer Interfaces (BCIs) that adapt to user engagement and support personalized learning. Electroencephalography (EEG) offers a non-invasive, cost-effective modality for capturing neural activity, though traditional methods often struggle with cross-subject variability and task-specific preprocessing. We propose leveraging Brain Foundation Models (BFMs), large pre-trained neural networks, to extract generalizable EEG features for cognitive load estimation. We adapt BFMs for long-term EEG monitoring and show that fine-tuning a small subset of layers yields improved accuracy over the state-of-the-art. Despite their scale, BFMs allow for real-time inference with a longer context window. To address often-overlooked interpretability challenges, we apply Partition SHAP (SHapley Additive exPlanations) to quantify feature importance. Our findings reveal consistent emphasis on prefrontal regions linked to cognitive control, while longitudinal trends suggest learning progression. These results position BFMs as efficient and interpretable tools for continuous cognitive load monitoring in real-world BCIs.

Problem

• Cognitive Load Estimation
• Brain-Computer Interfaces
• EEG
• Cross-subject Variability
• Real-time Monitoring

Innovation

• Brain Foundation Models
• Cognitive Load Estimation
• EEG
• Interpretability
• Partition SHAP