Suppression of interference and artifacts by the Signal Space Separation Method

Area of Science:

• Biophysics
• Neuroimaging
• Signal Processing

Background:

• Multichannel measurements capture complex vector fields, like magnetic fields from brain activity.
• Laplace's equation constrains magnetic fields, leading to spatially band-limited signals outside source volumes.
• Distinguishing neural signals from external interference is crucial for accurate analysis.

Purpose of the Study:

• To introduce and validate Signal Space Separation (SSS) as a method for removing external magnetic interference.
• To demonstrate the efficacy of SSS in isolating neural signals from environmental noise.
• To explore the application of SSS in transforming sensor configurations and compensating for movement artifacts.

Main Methods:

• Utilizing Laplace's equation to derive a functional solution for the magnetic scalar potential.
• Calculating individual basis vectors for functional expansion to create a signal basis.
• Performing unique SSS decomposition of signal vectors into source and interference components.

Main Results:

• SSS provides a unique decomposition of signal vectors, separating internal neural signals from external magnetic interference.
• The method is device-independent, allowing for transformation to virtual sensor configurations.
• SSS effectively compensates for distortions caused by object movement and enables recording of physiological DC phenomena.

Conclusions:

• Signal Space Separation is an elegant and effective method for removing external disturbances in multichannel magnetic field measurements.
• SSS preserves the morphology and signal-to-noise ratio of neural signals, enhancing the reliability of neuroimaging data.
• The device-independent nature of SSS offers flexibility in sensor configuration and artifact correction for advanced brain activity studies.