Looking through the windows: a study about the dependency of phase-coupling estimates on the data length

Area of Science:

• Neuroscience
• Computational Neuroscience
• Signal Processing

Background:

• Real-time characterization of functional connectivity (FC) dynamics is crucial for brain-computer interfaces and neurofeedback.
• Accurate FC estimation relies on detecting statistical dependencies in short data windows.
• Phase-coupling (PC) of neuronal oscillations is a key functional neural coupling mode.

Purpose of the Study:

• To evaluate the reliability of FC estimation based on data length using realistic simulations.
• To determine the optimal data window size for accurate PC estimation.
• To compare different PC metrics and spectral analysis approaches.

Main Methods:

• Generated synthetic data with varying signal-to-noise ratios, data lengths, and spectral analysis methods (Hilbert, Fourier).
• Compared seven PC metrics, including phase locking value (PLV) and phase lag index (PLI).
• Assessed the performance of different approaches under various conditions.

Main Results:

• Reliable PC estimates require 5-8 cycles of the target oscillation (e.g., 500-800ms at 10Hz) for signal-to-noise ratios >= 10 dB.
• Hilbert-based spectral analysis generally showed higher performance than Fourier-based methods.
• Narrow frequency band analysis necessitates a larger data window.

Conclusions:

• Established data length requirements for reliable real-time frequency-specific PC assessment.
• Findings support the development of best-practice guidelines for FC analysis.
• Informs the design of real-time neuroimaging and brain-computer interface systems.