Electroencephalography-Based Brain-Computer Interface System Using Tongue Movement Imagery for Wheelchair Control.

Area of Science:

• Neuroscience
• Assistive Technology
• Biomedical Engineering

Background:

• Brain-computer interfaces (BCIs) are crucial for restoring mobility in individuals with motor impairments.
• Existing tongue-controlled systems often require physical movements or intraoral devices, limiting usability.
• There is a need for more intuitive and comfortable BCI control methods.

Purpose of the Study:

• To introduce an electroencephalography (EEG)-based tongue motor imagery (MI) BCI for intuitive mental wheelchair control.
• To leverage preserved tongue motor function for natural, four-directional control.
• To evaluate the feasibility of imagined tongue movements for BCI applications.

Main Methods:

• EEG data were collected from 15 healthy participants using a 14-channel EMOTIV EPOC X headset.
• Six imagined tongue actions were designed to elicit alpha-band event-related desynchronization (ERD) over the tongue motor cortex.
• Extracted alpha-band ERD features were classified using various machine learning algorithms, including artificial neural networks (ANNs).

Main Results:

• The EEG-based tongue MI BCI demonstrated the potential for intuitive wheelchair control.
• Two-class tasks achieved an accuracy of 76.19%, with performance decreasing as task complexity increased.
• Artificial neural networks showed superior performance in multi-class scenarios.

Conclusions:

• The proposed tongue motor imagery BCI offers a promising approach for developing novel assistive technologies.
• Further research is needed to enhance classification techniques, user training, and real-time validation for practical usability.
• This method provides initial support for BCI control strategies in assistive technology.