Related Experiment Video
Updated: Jun 26, 2026
07:37
Assessment and Communication for People with Disorders of Consciousness
Published on: August 1, 2017
Visual P300-based BCI to steer a wheelchair: a Bayesian approach
Gabriel Pires1, Miguel Castelo-Branco, Urbano Nunes
1Institute for Systems and Robotics, University of Coimbra, 3000, Portugal. gpires@isr.uc.pt
Summary
This study introduces a novel brain-computer interface (BCI) using P300 detection for wheelchair control. The system achieved a 7 commands/min transfer rate with 95% accuracy, enhancing assistive technology.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Human-Computer Interaction
Background:
- Brain-computer interfaces (BCIs) offer potential for assistive technologies.
- P300 event-related potentials are a common BCI signal.
- Wheelchair control via BCI requires robust and accurate command selection.
Purpose of the Study:
- To develop and validate a new P300 paradigm for brain-computer interface-controlled wheelchair steering.
- To improve the accuracy and efficiency of BCI command selection for mobility assistance.
Main Methods:
- Utilized a novel P300 paradigm with visual stimuli (8 intensified arrows) for direction target selection.
- Employed a Bayesian approach for classification, incorporating prior statistical knowledge.
- Implemented Bayesian sensor fusion to combine multi-channel brain activity and event grouping for enhanced detection.
- System performance was adaptive to user and P300 pattern quality.
- Algorithms were trained offline and validated offline and online.
Main Results:
- Achieved a communication transfer rate of 7 commands/min.
- Attained 95% false positive classification accuracy.
- Demonstrated adaptive performance for improved user interaction.
Conclusions:
- The developed P300 paradigm is effective for BCI-controlled wheelchair steering.
- Bayesian classification and sensor fusion enhance BCI system robustness and accuracy.
- Adaptive system design improves usability and performance in assistive applications.
