Estimating Cognitive Effort from Functional Near-Infrared Spectroscopy (fNIRS) Signals using Machine Learning

2025-07-18

AI Summary

This study quantifies students’ cognitive effort during educational gaming to enable dynamic adaptation of instructional materials. Using functional near-infrared spectroscopy (fNIRS), we measured prefrontal cortical oxygenated hemoglobin signals, extracted temporal statistical and functional connectivity features, and applied multi-model machine learning to predict behavioral performance (accuracy: 58–67%). Our key contribution is the novel dual-metric framework—“relative neural efficiency” and “relative neural engagement”—which jointly integrates neural activation magnitude with behavioral output to robustly characterize trends in cognitive effort. Although single-trial prediction accuracy remains modest, both metrics exhibit strong consistency with task difficulty and learning phase progression. Results demonstrate that this paradigm enables objective, continuous monitoring of cognitive load dynamics. It thus provides an interpretable, deployable neurophysiological foundation for adaptive educational systems grounded in real-time neural feedback.

Technology Category

• Cognitive Modeling & Cognitive Systems: Adaptive Behavior
• Machine Learning: Neuro-Symbolic Learning
• Game Theory and Economic Paradigms: Adversarial Learning

Application Category

• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: Social networks and social learning

Abstract

The estimation of cognitive effort could potentially help educators to modify material to enhance learning effectiveness and student engagement. Where cognitive load refers how much work the brain is doing while someone is learning or doing a task cognitive effort consider both load and behavioral performance. Cognitive effort can be captured by measuring oxygen flow and behavioral performance during a task. This study infers cognitive effort metrics using machine learning models based on oxygenated hemoglobin collected by using functional near-infrared spectroscopy from the prefrontal cortex during an educational gameplay. In our study, sixteen participants responded to sixteen questions in an in-house Unity-based educational game. The quiz was divided into two sessions, each session consisting of two task segments. We extracted temporal statistical and functional connectivity features from collected oxygenated hemoglobin and analyzed their correlation with quiz performance. We trained multiple machine learning models to predict quiz performance from oxygenated hemoglobin features and achieved accuracies ranging from 58% to 67% accuracy. These predictions were used to calculate cognitive effort via relative neural involvement and efficiency, which consider both brain activation and behavioral performance. Although quiz score predictions achieved moderate accuracy, the derived relative neural efficiency and involvement values remained robust. Since both metrics are based on the relative positions of standardized brain activation and performance scores, even small misclassifications in predicted scores preserved the overall cognitive effort trends observed during gameplay.

Problem

• Estimating cognitive effort using fNIRS and machine learning
• Predicting quiz performance from oxygenated hemoglobin features
• Calculating neural efficiency and involvement during educational gameplay

Innovation

• Machine learning models predict cognitive effort
• fNIRS measures oxygenated hemoglobin in prefrontal cortex
• Cognitive effort derived from neural and performance metrics