Multiple window time-frequency distribution and coherence of EEG using Slepian sequences and hermite functions

Area of Science:

• Signal Processing
• Neuroscience
• Biomedical Engineering

Background:

• Nonstationary signals like electroencephalogram (EEG) require advanced analysis techniques.
• Conventional spectrograms may lack the precision needed for detailed EEG analysis.
• Multiple Window Time-Frequency Analysis (MW-TFA) is a novel approach for improved spectral estimation.

Purpose of the Study:

• To apply and evaluate MW-TFA techniques for electroencephalogram (EEG) signal analysis.
• To compare the performance of MW-TFA against conventional spectrogram methods.
• To assess the utility of MW-TFA for time-frequency (TF) coherence estimation in EEG.

Main Methods:

• Application of Multiple Window Time-Frequency Analysis (MW-TFA) to EEG data.
• Comparison of MW-TFA results with conventional spectrogram analysis.
• Evaluation of MW-TFA using Slepian sequences and Hermite functions as window sets.

Main Results:

• MW-TFA provides low bias and variance time-frequency distributions for EEG.
• MW-TFA enables time-frequency coherence estimation between simultaneously recorded EEG sites.
• No significant difference in TF representations was observed when using matched Slepian or Hermite windows.

Conclusions:

• MW-TFA is a valuable technique for detailed analysis of nonstationary EEG signals.
• The method offers superior performance over conventional spectrograms for EEG.
• MW-TFA facilitates robust estimation of spectral properties and inter-site coherence in EEG data.