On variability in local field potentials.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Signal Processing

Background:

• Neuronal coding and decoding depend on response stability; high variability can be detrimental.
• The Fano factor, a measure of spike count variability, decreases with sensory stimulation.
• This variability concept has been extended to electroencephalography (EEG) and local field potential (LFP) signals.

Purpose of the Study:

• To investigate the relationship between EEG/LFP variability and power.
• To propose a new metric for quantifying LFP power variability.
• To explore the implications of this metric for understanding neuronal coding and decoding.

Main Methods:

• Empirical analysis of EEG and LFP signals, focusing on intra-trial variance (power).
• Quantification of LFP power variability using the coefficient of variation (CV) of the log power ratio between active and baseline conditions.
• Examination of changes in CV(log(power ratio)) for gamma and alpha frequency bands under stimulation.

Main Results:

• A strong correlation was found between EEG/LFP variability and intra-trial variance (power), outside of evoked potentials.
• The CV(log(power ratio)) decreased when gamma and alpha power increased due to stimulation.
• The CV(log(power ratio)) increased when alpha power decreased due to stimulation, suggesting an inverse relationship.

Conclusions:

• LFP power, not the raw signal, is a more relevant measure for assessing signal variability.
• The proposed CV(log(power ratio)) metric effectively captures frequency-specific variability in neural signals.
• This metric offers a valuable tool for analyzing existing and future EEG, LFP, and MEG datasets for insights into neuronal coding.