EEG-EMG coupling as a hybrid method for steering detection in car driving settings

Area of Science:

• Neuroscience
• Human-Computer Interaction
• Biomedical Engineering

Background:

• Understanding mental processes in human behavior is crucial for developing advanced assistive driving technologies.
• Current assistive systems require reliable methods to interpret driver intentions, such as steering direction.

Purpose of the Study:

• To propose and validate a hybrid method using electroencephalography (EEG) and electromyography (EMG) to distinguish between left and right steering actions.
• To enhance the reliability and predictive power of driver monitoring systems for assistive driving.

Main Methods:

• Recorded 128-channel EEG and EMG from deltoids and forearm extensors in 24 participants during non-ecological and ecological steering tasks.
• Identified EEG mu rhythm modulation in non-ecological tasks and correlated it with EMG activity during ecological steering.
• Employed cross-correlation analysis to link non-ecological EEG features with ecological EMG signals for steering side detection.

Main Results:

• EEG mu rhythm de-synchronization in a non-ecological task was found to correlate with steering actions in an ecological driving scenario.
• The hybrid EEG-EMG system demonstrated significant cross-correlation, coupling brain and muscle activity.
• This hybrid approach detected the steering side earlier than EMG signals alone, overcoming limitations of ecological EEG data.

Conclusions:

• A hybrid system combining EEG and EMG provides a reliable and predictive method for detecting steering side in driving scenarios.
• This approach enhances the accuracy and adaptability of assistive driving devices by integrating cerebral and muscular signal characteristics.
• The findings suggest the potential for complementary physiological signals to control driver assistance levels effectively.