Fluctuating Physiological Noise in Cortical Activity During Visual Stimulation.

Area of Science:

• Biomedical Signal Processing
• Information Theory
• Neuroscience

Background:

• Physiological noise complicates analysis of complex biomedical signals.
• Non-stationary noise dynamics are challenging to quantify in electroencephalographic (EEG) data.
• Existing methods may not adequately capture time-resolved noise fluctuations.

Purpose of the Study:

• To develop and validate a novel method for time-resolved identification of non-stationary physiological noise.
• To apply this method to electroencephalographic (EEG) data.
• To investigate the dynamics of physiological noise during resting-state and visual stimulation.

Main Methods:

• Modeled physiological noise as a dynamical recursive realization of independent and identically distributed (IID) Gaussian random variables.
• Utilized Approximate Entropy, an information-theoretic quantifier, to estimate noise power.
• Employed a sliding window approach for time-resolved analysis.
• Validated the method on synthetic time series with controlled noise levels.

Main Results:

• The proposed method accurately estimated noise power, even with short window lengths.
• Significant time-resolved fluctuations in physiological noise were observed in real EEG recordings.
• A notable reduction in physiological noise levels occurred during visual stimulation in both occipital and frontal cortical regions.
• Findings suggest decreased noise complexity correlates with increased dynamic activity in relevant brain regions.

Conclusions:

• The developed method robustly identifies non-stationary physiological noise in EEG signals.
• The findings provide insights into physiological noise dynamics during cognitive tasks.
• This approach has potential implications for improving cortical signal analysis and clinical applications, possibly using noise as a biomarker.