Online Recursive ICA Algorithm Used for Motor Imagery EEG Signal.

Area of Science:

• Neuroscience
• Signal Processing
• Biomedical Engineering

Background:

• Electroencephalogram (EEG) signals are crucial for understanding brain activity.
• Independent Component Analysis (ICA) is a key technique for separating EEG sources from artifacts.
• Traditional ICA methods often assume instantaneous and offline processing, limiting real-time applications.

Purpose of the Study:

• To propose a novel framework using an extended online recursive ICA algorithm (ORICA) for motor imagery (MI) EEG analysis.
• To demonstrate ORICA's effectiveness in accurate and real-time source separation for artifact-contaminated MI EEG.
• To utilize topographic maps for identifying electrooculogram (EOG) signals within the separated components.

Main Methods:

• Extension of the online recursive ICA algorithm (ORICA).
• Application of the ORICA framework to motor imagery (MI) EEG recordings.
• Employing topographic maps to distinguish target signals from artifacts, specifically EOG signals.

Main Results:

• ORICA demonstrated adaptability for accurate and effective source separation in artifact-contaminated MI EEG.
• The proposed framework successfully identified EOG signals using topographic maps.
• Experimental results confirmed the framework's capability for real-time processing of MI EEG.

Conclusions:

• The novel ORICA-based framework is suitable for real-time processing of motor imagery EEG.
• The method effectively separates neural signals from artifacts, including EOG, in dynamic recordings.
• This advancement facilitates more robust and immediate analysis of brain activity during motor imagery tasks.