Summary Task 4 - MEG, EEG, oscillations & social cognition

TASK 4: MEG, EEG, OSCILLATIONS &
SOCIAL COGNITION
INTRODUCTION INTO MAGNETOENCEPHALOGRAPHY (MEG)

HOW DOES MEG WORK?

 Synaptic activity leads to small magnetic fields  perpendicular to electrical current
 Non-invasive
 Recorded with neuromagnetometer – positioned around outside of the head
 Underlying electrical activity deduced by mathematical modelling
 MEG traces can be recorded & averaged over a series of trails to obtain event-related fields
(ERFs)
 Magnetic fields detected using superconducting quantum interference devices (SQUIDs)
 Placed at various points on the surface of the scalp
 Magnetic fields must be sampled over a range of locations  distribution of electrical
currents inside brain can be calculated accurately
 Magnetometers – cover whole scalp & complete magnetic-field pattern can be measured
simultaneously
 Newest one – 306 SQUIDs at 102 measuring sites
 Each of 102 sensors measures in x, y & z directions
 Immersed in liquid helium at -169°C & positioned close to the head
 Preventing contamination with fields from e.g., power lines  measurements inside a
room made from several layer of aluminium & mu-metal
 Iron & nickel – high magnetic permeability  external magnetic fields are “trapped” in it,
shielding room inside
 All magnetic materials are forbidden inside the shielded room
 Direction of magnetic flux outside head determined by direction of current within
group of neurons , according to right-hand rule of electromagnetism

CALCULATING THE SOURCE OF MAGNETIC FIELDS

 Inverse problem – brain is spherical & active areas can be adequately represented by
single / multiple current dipoles
 Computer makes initial guess to where dipoles might be & then calculates external
magnetic field that these dipoles would produce
 Compares computed field to measured field
 Repeats calculation with dipoles at different positions until calculated &
experimental results match
 2 or more regions of brain active – measured magnetic field depends on position & strength
of dipoles & extent to which neurons in different regions fire at the same time

,  Minimum current estimate technique
 Gives most probable distribution of currents in brain, calculated according to
concept of minimum norm
 Advantage: used without making any specific assumptions about way in which
currents are distributed

ADVANTAGES & LIMITATIONS

 Same temporal resolution as ERPs BUT better spatial resolution
 Magnetic fields are not distorted as they pass through the brain, skull, scalp
 Limitations
 Current flow needs to be parallel to surface of the skull (recorded neurons usually
within sulci)
 Magnetic fields generated by brain are extremely weak  room that is
magnetically shielded from all external magnetic fields
 Room for improvement in signal-to-noise ratio
 Increased by placing SQUID sensors closer to brain BUT difficult, because they have to be
at liquid-helium temperatures at all times

MEG VS. EEG


MEG EEG


Based on dipolar currents, measure the same neuronal currents


Very good temporal resolution


Better spatial resolution – skull & scalp do Skull & scalp distort electrical potential
not distort magnetic fields


Currents have to be tangential to brain Better at detecting currents that originate
surface – all other currents cancel each deep inside the brain / are radially
other out oriented


Cheaper


DATA ACQUISITION & SIGNAL ANALYSIS IN EEG

 Important to adhere to standardised electrode locations – distance of 2-3cm between
electrodes required
 Improved special resolution with high-density recordings (64-128 electrodes
enough)