Estimating cognitive workload using a commercial in-ear EEG headset

Area of Science:

• Neuroscience
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Brain-Computer Interfaces (BCIs) are limited in real-world applications due to practical constraints.
• In-ear electroencephalography (EEG) offers a portable solution for BCI applications outside laboratory settings.
• Estimating mental workload is crucial for adaptive systems and understanding cognitive states.

Purpose of the Study:

• To evaluate a commercial in-ear EEG system (IDUN 'Guardian') for mental workload estimation.
• To compare the performance of in-ear EEG against conventional EEG.
• To investigate the role of high-frequency gamma band activity in workload estimation.

Main Methods:

• Participants performed n-back and mental arithmetic tasks at increased complexity.
• Simultaneous in-ear and conventional EEG data were collected.
• EEG signals were analyzed in low (1-35 Hz) and high (1-100 Hz) frequency ranges, with emphasis on gamma band activity.

Main Results:

• In-ear EEG achieved significant workload estimation, outperforming chance (44.1% for 4 classes, 68.4% for 2 classes in n-back).
• Conventional EEG showed higher accuracy (80.3%, 92.9%) than in-ear EEG.
• Surrogate in-ear measures improved accuracy (57.5%, 85.5%), and high-frequency gamma band activity enhanced workload estimation.

Conclusions:

• Commercial in-ear EEG demonstrates feasibility for mental workload estimation, albeit with lower performance than conventional EEG.
• High-frequency gamma band activity is a significant factor for improving workload estimation accuracy.
• Findings provide guidelines for enhancing future in-ear EEG systems for practical BCI applications.