Examples of cancellation of EEG and MEG signals. (A) Simulated data generated by two current dipoles in the primary visual cortex of the left and right hemispheres. The maps depict spatial patterns of signals over the occipital region when each dipole is active alone (LH: left hemisphere, RH: right hemisphere), and when the two dipoles are active simultaneously. Top: EEG isocontour maps; red and blue indicate positive and negative scalp potentials, respectively. Middle: MEG magnetometer isocontour maps; red and blue indicate positive and negative radial component of the magnetic field. Bottom: MEG gradiometer signals; each arrow depicts the signal in a pair of planar gradiometers. The arrows have been rotated by 90 degrees from orientation of the actual gradient vectors, to match the orientation of the arrows with the direction of putative underlying source currents. The locations (green dots) and orientations (green lines) of the dipoles are indicated in coronal and sagittal MRI slices. This example illustrates how cancellation can occur in all three types of sensor arrays, when sources of opposite orientation are close to each other. (B) Maps of MEG magnetometer signals for bilateral inferior temporal sources. For this particular source configuration, the cancellation is specific to mid-occipital MEG magnetometer sensors.

Cancellation index I_C for randomly distributed sources. The index was computed for an array of 60 EEG electrodes, 102 MEG magnetometers, and 204 MEG gradiometers. The mean and standard deviation over 1000 random selections of n dipoles are shown. For clarity, the standard deviation bars are plotted one-sided only.

Examples of EEG and MEG signals for extended patches of uniform cortical activity. A patch with a 4-mm (A) and 20-mm (B) radius in the left parietal cortex is shown in red on a cortical surface representation reconstructed from anatomical MRI. The dots and squares indicate the locations of the EEG scalp electrodes and the MEG triple-sensor units (one magnetometer and two gradiometers), respectively. Cancellation is illustrated by the signal per unit source element, obtained by plotting the maps for the 20-mm patch normalized by the ratio of the areas of the two patches (C).

Cancellation index I_C for cortical patches. The mean and standard deviation over 200 randomly located simulated patches of uniform activity for each of the different radii shown.

Comparison of the cancellation index I_C and the orientation non-uniformity measure I_O for cortical patches. Data from simulated patches with different radii up to 50 mm are shown for EEG (A), MEG magnetometers (B), and MEG gradiometers (C).