A Hybrid Brain-Computer Interface for Real-Life Meal-Assist Robot Control

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Existing meal-assist robots have non-intuitive interfaces, limiting usability for individuals with disabilities and the elderly.
• There is a need for more natural and user-friendly assistive devices for daily activities like eating.

Purpose of the Study:

• To develop a hybrid brain-computer interface (BCI) for a meal-assist robot system.
• To enable intuitive control of meal-assist robots using electroencephalography (EEG) signals.
• To enhance the independence and quality of life for individuals with disabilities and the elderly.

Main Methods:

• A hybrid BCI system was developed, integrating electroencephalography (EEG) signals measured via scalp electrodes.
• The system utilizes triple eye-blinks (EBs) for cycle initiation, steady-state visual evoked potentials (SSVEPs) for food selection, and electromyograms (EMGs) from chewing for cycle completion.
• Experiments were conducted on five subjects to evaluate system performance.

Main Results:

• High accuracy was achieved: 94.67% for EBs, 83.33% for SSVEPs, and 97.33% for EMGs.
• The information transfer rate (ITR) for SSVEPs was 20.41 bits/min.
• Low false positive rates (FPR) were recorded: 0.11/min for EBs and 0.08/min for EMGs.

Conclusions:

• The developed hybrid BCI system is feasible for controlling meal-assist robots.
• The system allows users to eat more naturally, increasing self-esteem.
• This technology has the potential to significantly improve the quality of life for disabled and elderly individuals.