Real-time control of a video game with a direct brain--computer interface

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Technology

Background:

• Direct brain-computer interfaces (BCIs) offer potential for controlling assistive devices.
• The LF-ASD (Left-Foot-Stimulus-Detection) EEG-based brain switch has shown promise in offline studies.
• Limited data exists on the LF-ASD's real-world usability with interactive applications.

Purpose of the Study:

• To evaluate the real-time usability of the LF-ASD brain switch.
• To compare the performance of able-bodied individuals and those with high-level spinal cord injury using the LF-ASD.
• To assess the LF-ASD's effectiveness in controlling a video game environment.

Main Methods:

• Able-bodied participants and individuals with high-level spinal cord injury used the LF-ASD brain switch.
• Participants controlled a video game in real-time over three 1-hour sessions.
• Switch activation rates and false-positive rates were recorded and analyzed.

Main Results:

• Both subject groups achieved switch activation rates between 30% and 78%.
• False-positive rates ranged from 0.5% to 2.2% across all participants.
• Switch classification accuracy exceeded 94% for all subjects, indicating high performance.

Conclusions:

• The LF-ASD brain switch demonstrates effective real-time control capabilities for individuals with spinal cord injuries.
• Performance in a real-time gaming environment was comparable between able-bodied users and those with spinal cord injuries.
• These findings support the potential of EEG-based BCIs for practical assistive technology applications.