Analysis of single trial movement-related brain macropotential

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Movement-related brain macropotentials are crucial for understanding motor control.
• Accurate identification of these potentials in single trials is challenging due to noise.
• Existing methods often struggle with signal-to-noise ratio and artifact reduction.

Purpose of the Study:

• To introduce a parametric method for identifying movement-related brain macropotentials on a single trial basis.
• To utilize an autoregressive with exogenous inputs (ARX) algorithm for enhanced signal analysis.
• To effectively characterize and reduce noise sources like electroencephalography (EEG) and electrooculography (EOG) artifacts.

Main Methods:

• A parametric identification method based on the ARX algorithm was developed.
• Brain macropotentials were analyzed on a single trial basis.
• EEG and EOG signals were modeled as exogenous inputs to isolate brain activity.
• Simulations and experimental data from three subjects were used for validation.

Main Results:

• The ARX model significantly improved the signal-to-noise ratio in single trials.
• The model accurately identified the contributions of both signal and noise.
• Efficient reduction of ocular artifacts was achieved using the same algorithm.
• Movement-related brain macropotentials exhibited high trial-to-trial variability.

Conclusions:

• The proposed ARX-based method offers a robust approach for single-trial analysis of brain macropotentials.
• This technique effectively enhances signal quality and reduces artifacts, improving data reliability.
• Observed trial variability may reflect cognitive processes such as motor programming and error evaluation.