Classification of Electroencephalogram Signal for Developing Brain-Computer Interface Using Bioinspired Machine...

Area of Science:

• Neuroscience
• Rehabilitation Engineering
• Biomedical Engineering

Background:

• Brain-Computer Interface (BCI) technology translates neural signals into device commands, aiding individuals with disabilities.
• Electroencephalogram (EEG) signals are commonly utilized for BCI operation due to their accessibility.
• Age-related variations in neural signal characteristics may impact BCI system performance.

Purpose of the Study:

• To investigate the performance differences in a mobile robot navigation BCI task between young and adult age groups.
• To analyze the influence of age on electroencephalogram (EEG) signal features and classification accuracy.
• To evaluate the efficacy of a BCI system using band power features and a bioinspired neural network.

Main Methods:

• Recruited twenty subjects across two age groups (20-28 and 29-40 years).
• Utilized a three-electrode system to acquire EEG signals during a mobile robot navigation task.
• Employed band power features and a neural network architecture trained with a bioinspired algorithm for classification.

Main Results:

• The young adult group achieved a maximum classification performance of 94.66%, while the adult group achieved 94.18%.
• Online testing demonstrated higher average accuracy for the young group (94.00%) compared to the adult group (92.00%).
• A slight decrease in signal quality was observed in the adult group, correlating with age.

Conclusions:

• Age is a significant factor influencing BCI performance, with younger individuals demonstrating superior accuracy.
• The developed BCI system, utilizing band power features and a bioinspired neural network, shows promise for mobile robot navigation.
• Further research is warranted to optimize BCI systems for diverse age demographics and improve signal processing techniques.