Empirical mode decomposition for sleep-wake analysis in a corticothalamic neural field model

Area of Science:

• Neuroscience
• Computational Neuroscience
• Signal Processing

Background:

• Corticothalamic oscillations are vital for sleep-wake regulation but their mechanisms are unclear.
• Conventional spectral analysis struggles with nonlinear, non-stationary neural oscillations.
• Understanding these oscillations is key for diagnosing neurological and sleep disorders.

Purpose of the Study:

• To apply iterated masking empirical mode decomposition (itEMD) to analyze simulated electroencephalographic (EEG) signals.
• To characterize frequency components of neural oscillations during different vigilance states using a computational model.
• To link computational model findings to empirical EEG data for improved sleep studies.

Main Methods:

• Utilized a physiologically-based corticothalamic neural field model to simulate EEG signals.
• Applied iterated masking empirical mode decomposition (itEMD) for signal decomposition into intrinsic mode functions (IMFs).
• Analyzed IMFs to identify distinct oscillatory patterns associated with sleep and wake states.

Main Results:

• itEMD successfully decomposed simulated EEG signals into characteristic IMFs.
• Identified distinct oscillatory patterns: dominant low-frequency rhythms in sleep and broadband activity in wakefulness.
• Model-derived IMFs showed correspondence with empirical EEG features during sleep.

Conclusions:

• itEMD is a powerful tool for analyzing complex neural signals from computational models.
• Revealed insights into the dynamical structure of corticothalamic oscillations and their role in state transitions.
• Provides a framework for integrating computational modeling with empirical sleep research for clinical applications.