Mental Fatigue Classification with High-Density Diffuse Optical Tomography: A Feasibility Study.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Cognitive Science

Background:

• Mental fatigue impacts daily functioning and clinical conditions.
• Monitoring mental states requires accessible and reliable tools.
• Functional near-infrared spectroscopy (fNIRS) offers potential for non-invasive brain monitoring.

Purpose of the Study:

• To evaluate wearable High-Density Diffuse Optical Tomography (HD-DOT) for detecting mental fatigue.
• To differentiate between low and high mental fatigue states using fNIRS data.
• To develop and validate a robust classification model for real-time applications.

Main Methods:

• Utilized wearable HD-DOT for data collection in healthy participants.
• Processed fNIRS data through filtering, channel selection, and Uniform Manifold Approximation (UMAP) for dimensionality reduction.
• Employed Support Vector Machines (SVM) with various kernel functions for classification.
• Developed participant-specific models and a leave-one-participant-out cross-validation strategy.

Main Results:

• Achieved an average accuracy of approximately 90% for participant-specific fatigue classification.
• Demonstrated high performance metrics including F1 score and Area Under the Curve (AUC).
• Validated the model's generalizability and effectiveness in distinguishing fatigue states.

Conclusions:

• Wearable HD-DOT is a highly effective tool for assessing mental fatigue.
• The developed SVM model shows significant potential for real-time Brain-Computer Interface (BCI) applications.
• This approach offers a promising avenue for continuous mental state monitoring.