Related Experiment Video
Updated: Sep 20, 2025
10:07
Using Fiberless, Wearable fNIRS to Monitor Brain Activity in Real-world Cognitive Tasks
Published on: December 2, 2015
27.2K
Automatic Cognitive Fatigue Detection Using Wearable fNIRS and Machine Learning
Rui Varandas1,2, Rodrigo Lima3,4, Sergi Bermúdez I Badia3,4
1LIBPhys (Laboratory for Instrumentation, Biomedical Engineering and Radiation Physics), Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal.
Sensors (Basel, Switzerland)
|June 10, 2022
Summary
This study used wearable sensors and machine learning to detect cognitive fatigue via Brain-Computer Interfaces (BCI). While accurate for some, individual validation is needed for real-world learning applications.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Machine Learning
Background:
- Wearable sensors enable unobtrusive patient monitoring and data generation in healthcare.
- Brain-Computer Interfaces (BCI) leverage these sensors for continuous cognitive state monitoring.
- Cognitive fatigue significantly impacts performance and attention, necessitating detection for optimized learning.
Purpose of the Study:
- To develop and evaluate a BCI system for automatic detection of cognitive fatigue.
- To apply this system in educational settings to prompt timely breaks.
- To investigate the efficacy of machine learning algorithms in identifying fatigue states.
Main Methods:
- Utilized two functional near-infrared spectroscopy (fNIRS) wearable devices to collect neuroimaging data.
- Developed an experimental protocol involving a digital lesson and cognitive tasks to induce fatigue.
- Implemented and user-tuned machine learning models for fatigue classification.
Main Results:
- Achieved classification accuracy of approximately 70.91 ± 13.67% for detecting cognitive fatigue.
- Found that 'time on task' was not a primary factor in inducing or detecting fatigue.
- Demonstrated that the developed BCI methodology requires individual validation for consistent application.
Conclusions:
- The fNIRS-based BCI shows potential for cognitive fatigue detection but requires personalized calibration.
- Individual validation is crucial before deploying fatigue monitoring in real learning environments.
- Future research should incorporate additional physiological signals and human-computer interaction metrics.
