A hybrid eog-eeg brain-computer interface for virtual reality applications to improve wheelchair driving skills

This thesis project aims to develop a VR game to control a wheelchair avatar through brain signals. Individuals with neuromotor diseases may require assistive devices or brain-computer interfaces (BCI) to control a powered wheelchair. Unfortunately, such technology requires extensive training to teach users how to properly and safely control wheelchairs. By utilizing virtual reality (VR) systems, individuals with disabilities can learn how to control BCI technology and improve their driving skills. To achieve this goal, we first conducted a systematic literature review to identify the signals and paradigms already used to develop BCIs for powered wheelchair control. Based on this search, we developed a BCI driven by electroencephalography (EEG) and electrooculogram (EOG) signals, combining the motor imagery (MI) paradigm with eye blinking. We then conducted a pilot study to verify the possibility of combining VR head-mounted display (HMD) hardware and EEG head caps. Results showed signal interferences while the HMD was used, so the experiment was conducted with the semi-immersive VR system, playing the 3D game on a laptop. Lastly, we evaluated the developed EOG-EEG BCI system with an able-bodied individual who controlled the virtual wheelchair avatar through a zig-zag pathway ten times. The system achieved an overall accuracy of 95%, demonstrating the feasibility of the BCI to control the VR system. Future work will involve testing the BCI system with individuals with neuromotor disabilities to evaluate how driving skills can be improved through VR. Overall, this thesis project demonstrates the successful use of a hybrid BCI system to control a virtual wheelchair, providing a promising avenue for individuals with neuromotor diseases to regain their independence and improve their quality of life.