From research to clinic: A sensor reduction method for high-density EEG neurofeedback systems

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• High-density electroencephalography (EEG) systems offer detailed source-estimated neurofeedback (NF) signals.
• Clinical application of high-density EEG is limited by cost and complexity.
• Developing methods to translate these signals to lower-density systems is crucial for wider adoption.

Purpose of the Study:

• To accurately deliver a source-estimated neurofeedback (NF) signal developed on a 128-sensors EEG system on a reduced 32-sensors EEG system.
• To create reliable reduced-sensor EEG montages for clinical NF applications.

Main Methods:

• A linearly constrained minimum variance beamformer algorithm identified key sensors from a 128-sensor EEG system.
• Monte Carlo-based sampling generated numerous 32-sensor montages from the selected sensors.
• K-means clustering analyzed high-performing montages to optimize a reduced 32-sensor montage.

Main Results:

• Nearly 4500 high-performing reduced montages were identified through Monte Carlo sampling.
• A refined set of 32-sensor montages reproduced the 128-sensor NF signal with over 80% accuracy in 72% of the test population.
• The developed method demonstrated the feasibility of accurate NF signal reduction.

Conclusions:

• A Monte Carlo reduction method successfully created reliable reduced-sensor EEG montages.
• These optimized montages can accurately deliver neurofeedback signals in clinical settings.
• This work provides a translational pathway for using low-density EEG systems for high-density NF measures.