Characterization of spatiotemporal dynamics in EEG data during picture naming with optical flow patterns

Area of Science:

• Neuroscience
• Computational Neuroscience
• Signal Processing

Background:

• Neural oscillations are fundamental to brain function.
• Understanding their spatiotemporal dynamics is crucial for deciphering neural processing.
• Existing methods may not fully capture the complex wave-like behaviors.

Purpose of the Study:

• To investigate the spatiotemporal dynamics of neural oscillations.
• To classify oscillation patterns into standing waves and modulated waves.
• To characterize these dynamics using optical flow patterns and validate findings with real EEG data.

Main Methods:

• Analysis of electric potential from neural activity.
• Identification of oscillation types based on frequency and phase.
• Application of optical flow patterns (sources, sinks, spirals, saddles).
• Comparison of analytical/numerical solutions with electroencephalography (EEG) data from a picture-naming task.

Main Results:

• Two types of neural oscillation dynamics were identified: standing waves and out-of-phase/modulated waves.
• Sources and sinks patterns were found to co-locate, with saddles positioned between them.
• The number of saddle patterns correlated with the total count of other patterns.
• Real EEG data confirmed these spatial correlations: source/sink clusters overlapped significantly (~60%), while saddle clusters showed minimal overlap (<1%) with source/sink clusters.
• Saddle patterns constituted approximately 45% of all identified patterns.

Conclusions:

• Optical flow analysis effectively characterizes spatiotemporal dynamics of neural oscillations.
• The identified spatial relationships between sources, sinks, and saddles provide insights into neural network organization.
• Findings are robust across simulated and real EEG data, suggesting a generalizable model for neural wave dynamics.