Standardized Kalman filtering for dynamical source localization of concurrent subcortical and cortical brain activity

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Tracking brain activity is crucial for understanding neurological function.
• Existing methods like Kalman filtering (KF) have limitations, including depth bias.
• Spatiotemporal standardization is needed to improve tracking accuracy.

Purpose of the Study:

• Introduce standardized Kalman filtering (SKF) as a novel spatiotemporal method for brain activity tracking.
• Reduce the depth bias inherent in non-standardized Kalman filtering (KF).
• Evaluate SKF's performance against existing methods.

Main Methods:

• Describe the standardized KF methodology from a Bayesian perspective.
• Utilize a realistic simulation of somatosensory evoked potential (SEP) data.
• Validate the method using real SEP data and compare with sLORETA and non-standardized KF.

Main Results:

• SKF accurately localized and tracked cortical and subcortical SEP originators across various signal-to-noise ratios.
• Compared to sLORETA, SKF maintained subcortical originator distinction at lower SNRs.
• Non-standardized KF mislocalized cortical activity and failed to track subcortical origins.

Conclusions:

• SKF combines the estimation accuracy of sLORETA with the traceability of KF.
• SKF produces focal estimates for SEP originators.
• SKF is valuable for studying time-evolving brain activity and localizing deep sources without prior knowledge.