Toward a reliable, automated method of individual alpha frequency (IAF) quantification

Area of Science:

• Neuroscience
• Cognitive Science
• Electrophysiology

Background:

• Individual alpha frequency (IAF) is a key electrophysiological marker linked to cognitive function and intelligence.
• Current methods for estimating IAF, such as peak alpha frequency (PAF) and center of gravity (CoG), lack consensus and can be inconsistent.
• Reliable IAF estimation is crucial for defining individualized frequency bands and understanding cognitive differences.

Purpose of the Study:

• To describe and evaluate an automated strategy for deriving PAF and CoG using a Savitzky-Golay filter (SGF).
• To assess the performance of the SGF technique in both empirical and simulated electroencephalography (EEG) data.
• To provide a reliable and easily integrable tool for improving IAF research.

Main Methods:

• Resting-state EEG power spectra were smoothed using a Savitzky-Golay filter (SGF).
• Two prevalent IAF estimators, peak alpha frequency (PAF) and center of gravity (CoG), were derived from the smoothed spectra.
• The SGF method was evaluated using empirical data from 63 healthy adults and simulated EEG datasets.

Main Results:

• The SGF technique reliably extracted target alpha components, even in noisy spectral conditions.
• The SGF routine demonstrated superior performance compared to automated peak detection methods without spectral smoothing.
• Statistical properties of the SGF-derived PAF and CoG estimates were consistent with previous findings.

Conclusions:

• The automated SGF method offers a fast, open-source, and reliable approach for estimating IAF.
• This technique enhances the consistency and replicability of research utilizing peak alpha frequency (PAF) and center of gravity (CoG) measures.
• The SGF method is easily integrated into popular M/EEG toolboxes, facilitating its widespread adoption in neuroscience research.