A phase lag index hardware calculation for real-time electroencephalography studies.

Area of Science:

• Neuroscience
• Signal Processing
• Computer Engineering

Background:

• The phase lag index (PLI) is crucial for analyzing functional brain connectivity from electroencephalograms (EEGs).
• Current offline implementations of PLI are computationally intensive, limiting real-time neurofeedback applications.
• There is a need for efficient algorithms to enable online EEG analysis.

Purpose of the Study:

• To propose and validate a novel hardware architecture for real-time calculation of the phase lag index (PLI) in EEGs.
• To overcome the computational limitations of existing offline PLI methods.
• To facilitate the development of real-time neurofeedback systems.

Main Methods:

• Development of a real-time architecture for PLI calculation.
• Implementation on a 32-bit, 16-channel system using a Stratix IV GX FPGA operating at 188.32 MHz.
• Performance evaluation through simulations and comparison with MATLAB software.

Main Results:

• The synthesized FPGA system successfully calculated the PLI in real-time.
• The system achieved a computational speed at least 66 times faster than MATLAB.
• The mean square error was maintained below 5.72×10^-6, demonstrating high accuracy.

Conclusions:

• The proposed architecture enables efficient, real-time calculation of the phase lag index for EEG analysis.
• This advancement supports the development of advanced online neurofeedback systems.
• The hardware-based approach offers significant speed improvements over software-based methods.