Estimation of effective brain connectivity with dual Kalman filter and EEG source localization methods

Area of Science:

• Neuroscience
• Signal Processing
• Biomedical Engineering

Background:

• Effective connectivity is crucial for brain functional mapping using electroencephalography (EEG).
• Understanding the influence of active brain regions on others is essential for neurological research.

Purpose of the Study:

• To propose a new method for estimating effective connectivity in the brain using EEG.
• To simultaneously estimate brain source activity and their temporal dependencies.

Main Methods:

• Utilized standardized low-resolution brain electromagnetic tomography (sLORETA) for brain active region localization.
• Applied a multivariate autoregressive (MVAR) model to extracted brain sources.
• Employed a dual Kalman filter to estimate model parameters and effective connectivity.

Main Results:

• The proposed dual Kalman filter method successfully estimated effective connectivity and updated source activity.
• The method demonstrated robustness and sensitivity to noise in simulated EEG signals.
• Performance was validated against other methods, showing acceptable results with minimal mean square error on both simulated and real EEG data.

Conclusions:

• The novel dual Kalman filter approach offers a powerful tool for brain functional mapping via EEG.
• Simultaneous estimation of source activity and effective connectivity improves accuracy, especially under noisy conditions.
• This method provides a reliable means for analyzing complex brain network dynamics.