Wireless EEG with individualized channel layout enables efficient motor imagery training

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Technology

Background:

• Motor Imagery (MI) neurofeedback holds promise for various applications, but its translation to everyday environments is limited.
• User-friendly EEG systems are crucial for home-based or clinical neurofeedback protocols.
• Channel selection significantly impacts the performance of Brain-Computer Interfaces (BCIs).

Purpose of the Study:

• To investigate the effects of systematic motor imagery (MI) neurofeedback practice in an everyday environment.
• To compare two channel-reduction approaches for optimizing MI neurofeedback.
• To evaluate a user-friendly, portable EEG system for MI training.

Main Methods:

• Sixteen BCI novices underwent four days of MI training, imagining hand movements with feedback.
• Individualized, high-density EEG recordings identified the most informative bipolar channels on day one.
• Common Spatial Patterns (CSP) were used for channel selection, compared against Independent Component Analysis (ICA) and standard 10-20 systems.

Main Results:

• Initial online classification accuracy averaged 85.1%, with CSP-selected channels outperforming other methods.
• Online MI accuracy significantly increased from day 2 (69.1%) to day 4 (73.3%).
• Improved accuracy was primarily attributed to reduced ipsilateral sensorimotor rhythm desynchronization.

Conclusions:

• Systematic MI practice using a user-friendly EEG system in everyday settings promotes MI learning.
• The findings support the feasibility of efficient, home-based or hospital-based MI neurofeedback protocols.
• This research bridges the gap between lab-based studies and practical neurofeedback applications.