| |
Abstract:
Combining functional EEG and MEG data results in improved
spatial resolu-tion of neural source reconstructions. Quasi-radial
sources can be detected by EEG only, whereas quasi-tangential,
superficial sources exhibit better Signal-to-Noise-Ratios and
specificity in MEG studies. Merging both modalities needs a
transformation of the measured data to a common basis using their
statistical significances. The conductivity properties of the head
have to be known in EEG modeling, whereas in MEG these values
cancel. One can use MEG to calibrate the not very well determined
in vivo conductivities for EEG forward models.
The individual, anatomical MRI data are not only used for overlay
with the measured data, field maps, and merged EEG / MEG
reconstruction results, but also serve to set up more realistic
volume conductor models using the Boundary Element Method with
semi-automatically segmented compartments.
The cortical gray matter layer can be used to constrain the
solution space of equivalent dipole and current density
reconstructions. Comparisons of single modality and merged EEG /
MEG reconstructions are presented for spherical and realistic
volume conductor models with and without anatomical
constraints.
|
