Related Experiment Videos
Brain-computer interface (BCI) operation: signal and noise during early training sessions
Dennis J McFarland1, William A Sarnacki, Theresa M Vaughan
1Laboratory of Nervous System Disorders, New York State Department of Health, Wadsworth Center, P.O. Box 509, Empire State Plaza, Albany, NY 12201, USA. mcfarlan@wadsworth.org
Summary
Electromyogram (EMG) contamination is common in early brain-computer interface (BCI) training, especially for users struggling with EEG control. This artifact typically diminishes as users improve their electroencephalogram (EEG) control.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Rehabilitation Technology
Background:
- Brain-computer interfaces (BCIs) enable control via brain signals like mu and beta rhythms.
- Electromyogram (EMG) signals from muscles can contaminate electroencephalogram (EEG) recordings.
- Understanding artifact presence is crucial for accurate BCI training.
Purpose of the Study:
- To investigate the occurrence and characteristics of EMG contamination in novice BCI users.
- To analyze how EMG contamination impacts cursor control during initial training sessions.
- To differentiate between EEG-driven control and EMG-driven cursor movement.
Main Methods:
- EEG and EMG data were collected from 7 new users over 10 BCI training sessions.
- Users attempted to control cursor movement using mu (8-12 Hz) or beta (18-25 Hz) rhythm amplitude.
- Data analysis focused on identifying and characterizing EMG contamination in relation to user performance.
Main Results:
- Five out of seven users achieved successful EEG-based cursor control.
- EMG contamination was observed in early sessions for some users, often linked to errors.
- For users who did not achieve EEG control, EMG artifact initially assisted cursor movement and persisted.
- EMG contamination decreased as EEG control improved in successful users.
Conclusions:
- EMG contamination is a frequent issue in early BCI training, particularly for unsuccessful trials or users lacking EEG control.
- EMG artifacts can initially facilitate cursor movement but should be distinguished from true EEG control.
- Comprehensive spectral and topographical analyses are essential to identify and manage EMG contamination during BCI training.