EEG-derived estimators of present and future cognitive performance.

Area of Science:

• Neuroscience
• Cognitive Psychology
• Human Factors Engineering

Background:

• Previous electroencephalography (EEG) research on fatigue focused on concurrent cognitive performance.
• Limited understanding of EEG's predictive capability for future cognitive states exists.

Purpose of the Study:

• To investigate EEG's ability to assess fatigue's impact on present and future cognitive performance.
• To develop EEG-based models for predicting performance decrements during sustained attention.

Main Methods:

• Utilized a 20-minute 3-choice active vigilance task (3CVT) with 50 subjects.
• Developed EEG models correlating brain activity with a combined reaction time (RT) and accuracy metric.
• Employed cross-validation to ensure model stability and prevent overfitting.

Main Results:

• EEG models demonstrated strong correlations (up to 0.89) for present and future performance estimation.
• Models accurately predicted "unacceptable" performance (missed responses, slow RTs) throughout the task.
• EEG successfully predicted gross performance degradations 5-15 minutes in advance.

Conclusions:

• EEG is a viable tool for assessing and predicting fatigue-induced cognitive performance decline.
• This predictive capability can inform interventions to mitigate performance decrements in vigilance tasks.
• EEG offers a non-invasive method for monitoring cognitive states and anticipating future performance limitations.