A Real-Time Magnetoencephalography Brain-Computer Interface Using Interactive 3D Visualization and the Hadoop...

Area of Science:

• Neuroscience and Big Data Analytics
• Biomedical Engineering
• Health Informatics

Background:

• Human biology data is rapidly expanding, presenting significant opportunities for healthcare, particularly in treating brain-related conditions.
• Next-generation IT and sensory devices are generating vast amounts of patient data, necessitating advanced storage and analysis solutions.
• Brain-Computer Interfaces (BCIs) offer innovative ways to interact with technology using neural signals.

Purpose of the Study:

• To present an innovative Brain-Computer Interface (BCI) for interactive 3D visualization.
• To utilize the Hadoop Ecosystem for efficient data analysis and storage of complex biological data.
• To demonstrate the BCI's capability in distinguishing thought actions from magnetoencephalographic (MEG) brain signals.

Main Methods:

• Implementation of Bayesian factor analysis algorithms within the BCI to interpret MEG signals.
• Collection of MEG data from five subjects during specific tasks.
• Development of a driver to translate BCI signals into mouse and keyboard commands for real-time applications.
• Utilizing the Hadoop Ecosystem for storing and analyzing subject MEG brainwaves and performance data.

Main Results:

• Achieved 90% positive performance in distinguishing mid- and post-movement brain activity using MEG signals.
• Successfully demonstrated a flight visualization simulation controlled by user's thoughts (left/right turns).
• The BCI driver components are adaptable for integration into various software applications, replacing traditional input methods.

Conclusions:

• The presented BCI system, leveraging the Hadoop Ecosystem, effectively analyzes MEG data for distinguishing thought actions.
• The BCI technology shows significant potential for advancing interactive 3D visualization and real-time control in various applications.
• This approach provides a robust framework for managing and analyzing large-scale neurophysiological data for enhanced human-computer interaction.