From Max Sutherland's weblog: www.sutherlandsurvey.com
Brain-Imaging Methods
Introduction:
There is a great deal of psychological interpretation involved in understanding the meaning of
increased brain activation, i.e. in specifying what mental process is signified by an activation.
Most imaging studies report activations arising from the difference between two tasks.1
As
described by Robinson 2004 a typical experiment has the subject lying still for thirty seconds,
then performs some task for thirty seconds, then lies still for thirty seconds. For each brain area,
the signal during the task is compared to the signal at rest; those areas of the brain with stronger
signals during the task are presumed to be processing the information.
A very recent breakthrough holds out eventual promise in the future of detecting the activity of an
individual neuron. At this stage however the smallest brain area that can be represented (a
voxel) is the size of a grain of rice and contains maybe tens of thousands of neurons. There are
something like100 billion neurons in the typical brain but current fMRI resolution is only about
150,000 `voxels'. The changes in blood flow in a voxel indicate increased activity of not a single
neuron but a huge pod of tens of thousands of neurons.
PET (Positron Emission Tomography)
Developed in the mid- 1970s, PET was the first scanning method to give functional information
about the brain. Both PET and FMRI provide information about
neural activity in different brain regions as indicated by the level
of cerebral blood flow. (With FMRI, the magnetic consequences
of blood oxygenation are measured whereas PET measures
blood flow by first injecting people with radioactive water and
measuring changes in radiation. 2
)
FMRI (Functional Magnetic Resonance Imaging)
MRI (Magnetic Resonance Imaging) requires no radioactive
materials and produces images at a higher resolution than PET. MRI is like an X-ray, taken inside
a giant doughnut magnet. Researchers realised that instead of taking snapshots of what injuries
look like, they could use MRI machines to see which parts of the brain are doing specific tasks
(such as perception, language and memory) ­ hence the term `functional' MRI. It involves very
rapid scanning of the brain to see which areas of the brain become activated. When neural
activity increases and the blood oxygenation in a region increases, this changes its magnetic
properties. Increased neural action draws a bigger blood supply to support its work, which shows
up--millisecond by millisecond --on an fMRI scan as magnetic changes. So, what fMRI detects
is not neural activity directly but magnetic changes that are blood-oxygen level dependent
(BOLD). The method is non invasive so multiple scans can be done on the same subject.
PET Scanner
[Note: The blood flow changes between several hundred milliseconds and several seconds after
the neuronal activity. Since changes in blood flow are linearly related to the amount of neuronal
activity, it is theoretically possible to extrapolate back from the blood flow measurements to the
time-course of neuronal activity. However, differences in the hemodynamic time-constants across
individuals and across different cortical regions can limit the temporal resolution to about a
second.]
MEG (Magnetoencephalography)
MEG's big advantage is that it can measure activity in the brain extremely quickly - every 1/1000
of a second, which is similar to the rate at which the brain works - essentially 'the speed of
thought'. MEG is a very different brain scanning technique
but used for similar purposes. It is closely related to
electroencephalography or EEG, since they both try to
measure the same neuronal currents. Electrical currents in
the brain's neuronal circuitry give rise to very weak
magnetic fields that can be picked up by superconducting
detectors arranged around the outside of the head. People
sit under a magnet that is not quite as daunting as an MRI
scanner. The main disadvantages of MEG are that it is
more expensive and not as good as fMRI at localising
where precisely in the brain, activity is taking place.
MEG Scanner
ERP ­ Event Related Potentials
Event Related Potentials: Scalp electrode
measurement.
(also called Evoked Response Potentials)
Electrodes on the scalp measure voltage fluctuations
resulting from electrical activity in the brain. The
"baseline" activity is then averaged out, leaving just
the electrical responses evoked by each stimulus
presentation. The location of where the activity is
generated inside the brain has to be imputed
mathematically. (In animal studies and patients
undergoing brain surgery, another way to localize ERP
sources is to place electrodes inside the brain.)
SSPT (Steady State Probe Topography)
For monitoring activity during dynamic stimulus
sequences, such as TV commercials only MEG and SSPT have
the necessary temporal resolution. (FMRI and PET do not provide
fast enough recording other than for relatively static stimulus
presentations.) SSPT measures steady-state visually evoked
potentials (SSVEP) and records at the rate of 13 times per second
from 64 electrodes in a lightweight skull cap. (Subjects also wear
lightweight goggles.)
Event Related Potentials: Electrode cap method.
1
Gabrieli J. D. E. (1998) Cognitive Neuroscience of Human Memory. In: Annual Review of Psychology, Vol.
49. Annual Reviews Inc, Palo Alto.
2
Gabrieli J. D. E. (1998) Cognitive Neuroscience of Human Memory. In: Annual Review of Psychology, Vol.
49. Annual Reviews Inc, Palo Alto.
SSPT cap & goggles