Mental workload classification using convolutional neural networks based on fNIRS-derived prefrontal activity

Area of Science:

• Neuroscience
• Biomedical Engineering
• Cognitive Science

Background:

• Functional near-infrared spectroscopy (fNIRS) is a valuable tool for assessing brain activity during cognitive tasks.
• Its application in measuring mental workload, particularly in individuals with mild cognitive impairment (MCI), requires further investigation.
• Convolutional neural networks (CNNs) offer potential for analyzing complex neuroimaging data.

Purpose of the Study:

• To investigate the feasibility of using CNNs with fNIRS-derived signals to classify mental workload in individuals with MCI.
• To evaluate the performance of CNNs in distinguishing different levels of cognitive demand.

Main Methods:

• Utilized spatial activation maps from prefrontal cortex activity of 120 MCI subjects performing N-back tasks at three difficulty levels (0, 1, 2-back).
• Employed CNNs for classification and evaluated performance using 5 and 10-fold cross-validation.

Main Results:

• Increased N-back task difficulty correlated with decreased classification accuracy and increased prefrontal activity.
• Significant differences in accuracy and prefrontal activity were observed across task difficulty levels (p < 0.05).
• CNNs achieved classification accuracies ranging from 0.83 to 0.96 for distinguishing task difficulty levels.

Conclusions:

• fNIRS is a promising tool for measuring mental workload in older adults with MCI, even with cognitive decline.
• This study demonstrates the feasibility and effectiveness of CNNs in classifying mental workload based on fNIRS signals from the prefrontal cortex.