Spatial filtering and neocortical dynamics: estimates of EEG coherence

Abstract

The spatial statistics of scalp electroencephalogram (EEG) are usually presented as coherence in individual frequency bands. These coherences result both from correlations among neocortical sources and volume conduction through the tissues of the head. The scalp EEG is spatially low-pass filtered by the poorly conducting skull, introducing artificial correlation between the electrodes. A four concentric spheres (brain, CSF, skull, and scalp) model of the head and stochastic field theory are used here to derive an analytic estimate of the coherence at scalp electrodes due to volume conduction of uncorrelated source activity, predicting that electrodes within 10-12 cm can appear correlated. The surface Laplacian estimate of cortical surface potentials spatially bandpass filters the scalp potentials reducing this artificial coherence due to volume conduction. Examination of EEG data confirms that the coherence estimates from raw scalp potentials and Laplacians are sensitive to different spatial bandwidths and should be used in parallel in studies of neocortical dynamic function.

Steady-State Visual Evoked Potentials: Distributed Local Sources and Wave-Like Dynamics Are Sensitive to Flicker Frequency

“…The EEG was mathematically referenced to the average potential of the 110 channels. Although no EEG reference is "ideal", the average reference enjoys the advantage of some theoretical justification (Bertrand et al, 1985;Nunez and Srinivasan, 2005) and performs adequately as an approximation to reference-independent potentials in simulation studies of EEG coherence with identical 128 channel electrode arrays (Srinivasan et al, 1998).…”

“…The model consists of a brain sphere (r 1 = 8.0 cm), surrounding by 3 shells representing the CSF (r 2 = 8.1 cm), skull (r 3 = 8.6 cm), and scalp (r 4 = 9.2 cm). The details of the solution for dipole sources have been published previously (Srinivasan et al, 1998). This model has been used extensively in simulation studies of surface Laplacians (Nunez et al, 1994;Srinivasan et al, 1996), EEG coherence estimates (Nunez et al, 1997;Srinivasan et al, 1998), and forms the basis for a number of approaches to source localization (Lopes da Silva 2004).…”

“…The details of the solution for dipole sources have been published previously (Srinivasan et al, 1998). This model has been used extensively in simulation studies of surface Laplacians (Nunez et al, 1994;Srinivasan et al, 1996), EEG coherence estimates (Nunez et al, 1997;Srinivasan et al, 1998), and forms the basis for a number of approaches to source localization (Lopes da Silva 2004). The thickness (or radii) and conductivity ratios between the spherical shells are the essential features of the model.…”

“…Large dipole layers, e.g., of radius 5 cm, that make a large contribution to the potential make limited contributions to the surface Laplacian. This strong influence of source region size on scalp potentials has also been quantified as preferential sensitivity to low spatial frequency components of the source distribution (Srinivasan et al, 1998;Nunez and Srinivasan, 2005). The selective sensitivity of the surface Laplacian to smaller source regions is a consequence of its preferential sensitivity to higher spatial frequencies than the scalp potentials, but limited by the separation distances between electrodes and between sources and electrodes (Srinivasan et al, 1996).…”

“…A simulation study was carried out using the 4-sphere model of the head and a spatial white noise source distribution (Srinivasan et al, 1998). A hemi-spherical layer composed of ~ 3600 dipole sources each with a random time series was chosen as the source distribution used to simulate scalp potential at 129 sites (electrodes).…”