Hybrid EEG-fNIRS Asynchronous Brain-Computer Interface for Multiple Motor Tasks.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Non-invasive Brain-Computer Interfaces (BCI) offer potential for neuroprosthetics and assistive technologies.
• Combining Electroencephalography (EEG) and functional Near-Infrared Spectroscopy (fNIRS) can enhance BCI capabilities.
• Sensory Motor rhythm (SMR)-based BCIs are crucial for controlling external devices through brain signals.

Purpose of the Study:

• To investigate novel methods for combining EEG and fNIRS in an asynchronous SMR-based BCI.
• To address the inherent delay in fNIRS hemodynamic responses by introducing new signal features.
• To classify four distinct executed movements using the integrated EEG-fNIRS BCI system.

Main Methods:

• Implemented an asynchronous SMR-based BCI system integrating EEG and fNIRS data.
• Developed and applied novel features, including slope indicators, to mitigate fNIRS signal delay.
• Utilized Common Spatial Patterns (CSPs) on both EEG and fNIRS signals, regularized and optimized via genetic algorithms across 15 participants.
• Classified four movement tasks: Right-Arm, Left-Arm, Right-Hand, and Left-Hand.

Main Results:

• The proposed methods, incorporating slope indicators and regularized CSPs, effectively reduced the fNIRS signal delay.
• Dynamic accuracy analysis demonstrated improved real-time performance of the combined EEG-fNIRS BCI.
• The system successfully classified four different executed movements, showcasing the efficacy of the integrated approach.

Conclusions:

• The combination of EEG and fNIRS, with novel signal processing techniques, enhances the performance of asynchronous SMR-based BCIs.
• The introduced slope indicators and optimized CSPs significantly diminish the fNIRS hemodynamic delay, enabling faster and more accurate movement detection.
• This research contributes to the advancement of more responsive and effective neuroprosthetic and assistive devices.