Analysis of spontaneous magnetoencephalography data by similarity measures

We discuss application of dissimilarity measures technique to the classific ation of time series obtained in experimental measurements, and use it to a nalyze multi-channel magnetoencephalography data (MEG), Dissimilarity measu res for the 1/f component of MEG data allow to distinguish signals coming f rom left and right hemispheres. The alpha-wave component does not allow thi s distinction, supporting the earlier research, indicating that alpha-wave sources travel along the whole of the posterior region of the brain. (C) 19 99 Elsevier Science B.V. All rights reserved.