Results: in a magnetoencephalographic investigation of the auditory evoked
field (AEF) in 17 schizophrenics and 17 controls, 37% of the schizophrenics
and 12% of the controls showed eye artifacts in every second trial or even
more frequently. In the uncorrected average fields, the ratio between the
power of artifacts and the power of the magnetoencephalogram (MEG) exceeded
the value of 0.1 for 48% of the schizophrenics and for 29% of the controls
. Ocular artifacts biased the locations of equivalent current dipoles of th
e M100 component towards deeper positions. A regression algorithm for the c
orrection of ocular artifacts in raw data and an identification technique o
f ocular artifacts based on the topography of transmission coefficients is
described.
Conclusions: A linear dependence of ocular artifacts in AEF an the electroo
culogram (EOG) was confirmed. Possible errors introduced by the correction
are discussed. Transmission coefficients should be calculated for several i
ndividual trials with the same type of artifact. Errors due to evoked poten
tials in the EOG were found to be comparable in amplitude to noise in the A
EF. Examples of transmission coefficients from the EOG to the MEG are given
. (C) 1998 Elsevier Science Ireland Ltd. All rights reserved.