060629 Freeman

The Bernstein Center for Computational Neuroscience Freiburg

Announcement for the next

BCCN Seminar

Prof. Walter J Freeman

University of California, Berkeley
U.S.A.

Coexistence of local and global neural activity fields revealed by simultaneous MEG and scalp EEG at high temporal and spatial resolution

A comparison was undertaken of the spatiotemporal resolution of spatial and temporal textures in neocortical neural activity derived from multichannel MEG and EEG data simultaneously recorded from an awake human subject at rest with eyes closed or engaged in a simple visual cognitive task. MEG signals were recorded from 306 sensors at 102 sites spaced on average 3.4 cm apart, each site with a magnetometer and 2 planar gradiometers. The EEG signals were recorded from 70 scalp electrodes in a modified 10-20 array on average 2.6 cm apart. The shortest distance was calculated between the scalp and pia for each sensor in a structural MRI with fiduciary markings. The Hilbert transform gave the analytic phase and analytic amplitude for each signal. The time series, temporal spectra and amplitude histograms of MEG and EEG and their analytic signals did not differ significantly. The EEG yielded coordinated spikes and dips in analytic phase differences (CAPD) that were correlated with alpha activity (eyes closed) or theta activity (eyes open); the MEG from magnetometers and gradiometers did not. Statistically significant positive and negative correlations were found between amplitudes and sensor distances to pial surfaces of gyri and sulci, as predictable in accord with inverse square and inverse cube laws. We concluded that the distances between sensors and subpial cortical sources of electric and magnetic fields were as important as the orientation of dendritic current dipoles in determining the magnitudes of recorded EEGs and MEGs. Global MEG signals that oscillated in phase in both opposing faces of sulci tended to cancel above the scalp surface. Local MEG signals on either of the two faces were not subject to this limitation. Therefore the EEG was better suited for studies of widely synchronized beta and gamma activity, and MEG was better suited to record focal activity relating to cognition. For both signals the spatial resolution an interval of 3 mm between sensors would be optimal.

Host: Prof. Ad Aertsen