Advancing Wearable BCI: Headphone EEG for Cognitive Load Detection in Lab and Field
Michael Thomas Knierim, Christian Zimny, Gabriel Ivucic, Tobias Röddiger
Abstract
By tracking changes in brain activity, researchers are constantly working to revolutionise human-technology interaction. Unfortunately, such brain-computer interfaces still face limitations in terms of wearability, making large-scale data collection difficult. While gel-based wearable solutions exist, validated dry electrode systems are needed for practical everyday use. In this article, two experiments with 50 participants and 146 recordings were conducted in laboratory and field settings to compare the wearability and performance of two dry-electrode EEG systems: the Open ExG headphones and the OpenBCI Ultracortex full-head EEG. Our results show that the headphone EEG is perceived as more wearable, has equal signal quality and recording reliability when set up by a trained experimenter in the lab, and shows reliable performance in a relevant application scenario: classification of cognitive load levels across four tasks. Field evaluations further validate these results through reliable load monitoring across recording sessions, after self-setup by study participants at home. While some limitations remain for wider field use of dry-electrode headphone EEG, we highlight necessary and achievable improvements for future system and study designs for real-world use.