MEG & EEG » Modeling
Magnetoencelphlography (MEG) and Electroencephalography (EEG) is used
to image electrical activity in the brain. Clusters of thousands of synchronously
activated pyramidal cortical neurons are believed to be the main generators
of MEG and EEG signals.
Figure: Excitatory postsynaptic potentials are generated
at the apical dendritic tree of a cortical pyramidal cell and trigger
the generation of electrical current. Large cortical pyramidal nerve cells
are organized in macro-assemblies with their dendrites normally oriented
to the local cortical surface. Functional networks made of these cortical
cell assemblies and distributed at possibly mutliple brain locations are
thus the putative main generators of MEG and EEG signals.
Figure: MEG sensor arrangement (left); recent developments
include whole-head sensor arrays for the monitoring of brain magnetic
fields at typically ~300 locations. The sensor measurements (middle) are
used to reconstruct brain activity in the visual cortex (right). Click
image to enlarge.
Forward Solutions
To estimate the neural sources of scalp potentials and neuromagnetic
fields, we must also be able to solve the associated forward problem,
i.e. we need a forward model that maps a source of known location, strength,
and orientation to an array of EEG or MEG sensors.
Figure: Spherical head models are frequently used
in forward modeling. They considerably simplify the electromagnetic equations
while at the same time they give good solutions.
Figure: Numerical solutions to the forward problem
include Finite Element Methods (FEM) and Boundary Element Methods (BEM).
They provide solutions based on solving partial differential equations
for the potentials on the surfaces or volumes. Click image to enlarge.
Visualization
Figure: Visualizing brain activity on the original
cortical surface is difficult because of the many sulci and gyri. Smoothing
the brain surface permits visualization of deep areas like the corpus
callosum.
Inverse Solutions
Scanning and imaging are the two approaches that have been widely used
to model the position of MEG / EEG signal generators.
Figure: The imaging approaches are based on the assumption
that the primary sources are intracellular currents in the dendritic trunks
of cortical pyramidal neurons, that are aligned normally to the cortical
surface. Various approaches can produce reconstruction / detection maps
on the cortex, including Tikhonov Regularized min-norm, Multiple Signal
Classification (MUSIC), and Beamformers.
Figure: In scanning approaches, MEG sources can be
represented by a few, usually one or two, equivalent current dipoles of
unknown location, and moment to be estimated with a nonlinear numerical
method. The figure presents an example source on the right temporal lobe.
|
