GALOTTI, SOLSO

Fifty years ago, neuroscientists had only a few tools and techniques to use in the direct observation
and exploration of the human brain. These included ablations (damaging tissue by excising or
freezing), single-cell recordings, postmortem examinations, and animal studies. Psychologists, on the
other hand, invented a whole arsenal of techniques in which the mind revealed itself, such as the
momentary presentation of stimuli and measuring of reaction time. New instruments have been
invented that have profoundly accelerated our understanding of the brain and, for our purposes,
spun off a new breed of scientist who is part neuroscientist and part cognitive psychologist. The new
technology was originally developed for the diagnosis of brain disorders, but it has now become a
valuable research tool for scientists interested in studying the mind. These methods have led to
important new discoveries in the study of human cognition and have already proved to be an
integral part of the future of cognitive neuroscience. These methods to examine the functioning of
normal brains using noninvasive means are collectively known as brain imaging techniques

CT Scans

Some of these methods give us information about neuroanatomy—the structures of the brain. One
of the earliest such brain-imaging techniques, developed in the 1970s, was X-ray computed
tomography—also called X-ray CT, computerized axial tomography, or CAT scan—a technique in
which a highly focused beam of X-rays is passed through the body from many different angles.
Differing densities of body organs (including the brain) deflect the X-rays differently, allowing
visualization of the organ.

(CT) is the process of using computers to generate a threedimensional image showing the structure
of the brain from flat (in other words, twodimensional) X-ray pictures. The machine rotates around
the skull, bombarding it with thin, fan-shaped X-ray beams. The beams pass through the brain and
are recorded on sensitive detectors on the opposite side of their source. This procedure is different
from a conventional X-ray examination, which gives only one view of the body part. Also, with
conventional X-rays, large molecules (such as calcium in the skull) absorb the rays and partially
occlude the organs behind them. The CT scan rotates the X-ray beam 180 degrees, resulting in
numerous “pictures” of the same organ and producing an internal cross section, or “slice,” of the
body part. This graphic cross section, called a tomogram, has become critical in medical diagnosis.

Typically, CAT scans of a person’s brain result in 9 to 12 different “slices” of the brain, each one
taken at a different level of depth. CAT scans depend on the fact that structures of different density
show up differently. Bone, for example, is denser than blood, which is denser than brain tissue,
which is in turn denser than cerebrospinal fluid (Banich, 2004). Recent brain hemorrhages are
typically indicated by the presence of blood; older brain damage, by areas of cerebrospinal fluid.
Thus clinicians and researchers can use CAT scans to pinpoint areas of brain damage and also to
make inferences about the relative “age” of the injury.

An even more sophisticated version of the CT technique, the dynamic spatial reconstructor (DSR),
shows internal structures in three dimensions. One advantage of CT is the ubiquity of the machines.
As of the mid-1990s, more than 10,000 scanners were in use in American hospitals. Also, recent
technology has helped solve one of the problems with this technique. The temporal resolution, or
shutter speed, had been about 1 second, with the result that dynamic processes (even the
heartbeat) appeared blurred. Now, an ultrafast CT has been developed that speeds up processing so
that previously blurred images are now clear. Additionally, a new CT imaging technique is called x-
ray microtomography. This technology uses CT to scan via a microscope which allow for 3D images
of very small structures (a 5mm fossilized fish eyeball, the tine structures of the inner ear, and the
root structure of a human molar tooth (Uzon, Curthoys, & Jones, 2006). The full utility of this
technology is yet to be determined.

PET Scans

Positron emission tomography (PET) which dates back to the 1970s, involves injecting a radioactively
labelled compound (radioisotopes of carbon, nitrogen, oxygen, or fluorine subatomic particles that
rapidly emit gamma radiation, which can be detected by devices outside the head). PET scans
measure the blood flow to different regions of the brain, allowing an electronic reconstruction of a
picture of a brain, showing which areas are most active at a particular time (Posner & Raichle, 1994).
It is used to scan for glucose use in the brain. PET scans differ from CT scans in that they use
detectors to measure radioactive particles in the bloodstream to measure regional cerebral blood
flow (rCBF). Active parts of the brain require blood flow, and hence, more radioactive tracers amass
in operative areas. These tracers emit rays, which can be converted to visual maps. The computer
codes the glucose absorption data to show varying levels of activity in a color-coded map. Red
usually indicates more active areas and blue usually indicates less activate areas. The application of
PET scans to cognitive neuroscience has been particularly useful for measuring brain functioning. At
Lund University in Sweden, research scientists Jarl Risberg and David Ingva (see Lassen, Ingvar, &
Skinhoj, 1979), working in collaboration with Steve Petersen, Michael Posner, Marcus Raichle, and
Endel Tulving, have pioneered the use of PET scans in cognitive psychology (see Posner, Petersen,
Fox, & Raichle, 1988). PET technology is more expensive compared to CT scanners. Some advances in
PET studies have been in the type of tracers injected into the blood. Early studies were based on
inhaling a radioactive isotope. Nowadays, a different isotope is used that can be injected
intraveneously, which allows for high-resolution maps to be made in a few seconds (Risberg, 1987,
1989; Tulving, 1989a, 1989b), thus giving the researcher considerable latitude in collecting cognitive
data.

A variation of the PET procedure involves measuring local metabolic changes instead of blood flow,
using an injection of fluorodeoxyglucose, a radioisotope structurally similar to glucose. PET scans rely
on the fact that when an area of the brain is active more blood flows to it, and its cells take up more
glucose from the blood vessels that penetrate it (Frith & Friston, 1997; Kung, 1993). People
undergoing a PET brain scan sit with their head in a ring of photocells. A radioactive tracer, typically
15O2 (oxygen with one electron removed), is injected into a vein as water (that is, as H215O). Within
30 seconds, the tracer starts to reach the brain. The tracer 15O accumulates in the brain in direct
proportion to the amount of blood flowing to that brain region (Banich, 2004). Within the roughly
two minutes before the radioactive tracer decays to its half-life several scans can be made, showing
the amount of blood flowing to that region (Frith & Friston, 1997).

SPECT Scan

Another technique to measure cerebral blood flow is known as single-photon emission computed
tomography, or SPECT for short. The basic technique is similar to a PET scan, but does not require
some of the expensive equipment a PET scan does; thus it is sometimes known as a “poor person’s
PET.” Like CAT scans, however, PET and SPECT scans use radiation

MRI and fMRI

Magnetic resonance imaging (MRI) scans provide still images of structures of the brain. In the MRI
technique the body is surrounded with very powerful electromagnets that align the nuclei of
hydrogen atoms found in water. From these measures, it is possible to infer varying densities of
hydrogen atoms and their interaction with surrounding tissues. Since hydrogen reflects water
content, it is possible to use the MRI for diagnostic and research purposes. One of the main
drawbacks of the technique, until recently, was the time it took to form images using MRI
technology. Because it required long exposure time, the technique was acceptable for viewing static
biological structures. However, it was nearly useless for rapidly changing cognitive functions. It is
now possible using fMRI to apply high-performance image acquisition techniques to capture an
image in as little as 30 milliseconds, which is brief enough to record fast-paced cognitive functions.

Someone undergoing an MRI typically lies inside a tunnel-like structure that surrounds the person
with a strong magnetic field. Radio waves are directed at the head (or whatever body structure is
being scanned), causing the centres of hydrogen atoms in those structures to align themselves in
predictable ways. Computers collate information about how the atoms are aligning and produce a
composite three-dimensional image from which any desired cross-section can be examined further.

MRI scans are often the technique of choice, as they now produce “textbook-quality anatomy
pictures” of a brain (Carlson, 2004, p. 192). However, not everyone can undergo an MRI scan. The
magnetic fields generated in an MRI scan interfere with electrical fields, so people with pacemakers
are not candidates for an MRI (pacemakers generate electric signals). Nor are people with metal in
their bodies, such as a surgical clip on an artery or a metal shaving in the eye. The magnetic field
could dislodge the metal in the body, causing trauma.

Metal anchored to hard surfaces, such as dental fillings, is not a problem. Because MRIs require
people to lie very still in a tunnel-like machine that often leaves little room for arm movements,
people with claustrophobia are also not good candidates for this technique.

While CAT and MRI scans provide pictures of brain structures, and investigators can use these
pictures to pinpoint areas of damage or other abnormality, they provide only static pictures of the
parts of a brain. No information is gleaned about how a brain functions— that is, what areas of the
brain show activity when people perform different cognitive tasks. To answer such questions,
different brain-imaging techniques are needed. Fortunately, recent developments have created
techniques that fit the bill.

fMRI

(functional magnetic resonance imaging) detects increased blood flow to activated areas of the
brain, thus displaying function and structure. Functional magnetic resonance imaging, or fMRI, relies
on the fact that blood has magnetic properties. As blood is carried from the heart, it is maximally
magnetic. As it passes through capillaries, itbecomes less magnetic. Brain regions that are active
show a change in the ratio of oxygenated to deoxygenated blood (Banich, 2004). Such fMRI scans
use existing MRI equipment but provide clinicians and investigators with a noninvasive,
nonradioactive means of assessing blood flow to various brain regions.

TMS

Transcranial magnetic stimulation (TMS) is used in conjunction with EEG or MEG to evaluate the
effects of changes in brain electrical activity on perceiving and thinking. A magnetic charge is
directed via a wand placed on the head and pointed to a specific location in the brain for a very short
period of time. This charge alters the neural functioning; these effects on brain functioning can be
seen in the output of the EEG or MEG, as well as in terms of the person’s responses to the cognitive
and perceptual tasks he or she is engaged in at the time.

Micro CT
A new CT imaging technique is called x-ray microtomography. This technology uses CT to scan via a
microscope which allow for 3D images of very small structures (a 5mm fossilized fish eyeball, the
tine structures of the inner ear, and the root structure of a human molar tooth (Uzon, Curthoys, &
Jones, 2006). The full utility of this technology is yet to be determined.

EEG

Electroencephalography (EEG) can be used to detect different states of consciousness. Metal

MEG

Magnetoencephalography (MEG) uses a machine that measures brain activity from the outside of
the head by detecting the faint magnetic fields that brain activity produces. MEG produces an
activity map or functional image of the brain. Of all the brain scanning methods, MEG provides the
most accurate resolution of nerve cell activity (to the millisecond).

A newer technique, magnetoencephalography, or MEG, measures changes in magnetic fields

ERP

Another electrical recording technique, called event-related potential, or ERP, measures an area of
the brain’s response to a specific event. Participants in an ERP study have electrodes attached to
their scalp and are then presented with various external stimuli, such as sights or sounds. The
recording measures brain activity from the time before the stimulus is presented until some time
afterward. The brain waves recorded also have predictable parts, or components. That is, the shape
of the waveform can vary depending on whether the participant expects the stimulus to occur or is
attending to the location in which the stimulus appears, and whether the stimulus is physically
different from other recent stimuli.

Research Using Imaging Techniques:

Of particular interest to cognitive psychologists is the use of cortical blood flow patterns in memory
research. PET and fMRI measure blood flow. Endel Tulving has developed a theory of memory that
posits two unique types: episodic (memory for personal events) and semantic (memory for general
knowledge). In one experiment studying two types of memory by Tulving (1989b), participants were
asked to think silently about an event in their lives (an episodic memory) and then to think about
something general. The research was conducted with a PET scanner. Recall that PET scans work by
injecting a tracer into the participant’s bloodstream. While expertise is required to fully interpret
these scans, one can still see general differences in the patterns of blood flow, suggesting increased
neural activity associated with different regions of the brain. Basically, it appears that episodic
memory is identified by greater activation of the anterior portion of the cerebral cortex, and
semantic memory is associated with greater activation of the posterior regions. It seems safe to
conclude that episodic and semantic memory systems in- volved different brain processes and that
each has its own location. This in turn suggests that we may have multiple memory systems, an idea
long held in cognitive psychology. Data from this PET study supports other observations from clinical
studies of patients with brain damage who have episodic memory loss.
In another attempt to find a direct correlation between cognitive processes and brain activity,
Michael Posner, Steven Petersen, and their colleagues at the McDonnell Center for Higher Brain
Functions at Washington University conducted a series of PET experiments dealing with the
processing of words by the normal, healthy brain (Petersen, Fox, Posner, Mintun, & Raichle, 1988).
There were four stages of the experiment: (1) a resting stage; (2) the appearance of a single word on
a screen; (3) the reading of the word aloud; and (4) the production of a use for each word (see Figure
14). Each of these stages produced its own pattern of activation, allowing the researchers to localize
the different cognitive functions associated with seeing, speaking, and producing words.

This is also evidence for more holistic processing, with words in general being processed using vast
areas of the brain (i.e., there isn’t a little “word spot” in the brain). When a participant looked at a
word on a screen, the occipital region of the cortex was activated; when the participant was hearing
a word, the central part of the cortex was activated; when speaking a word, motor regions were
activated; and when asked to produce related words (e.g., if the word “cake” appeared, the
participant was to produce a verb related to that word, such as “eat”), the associative region
produced the greatest amount of activity, but other general activity throughout the cortex was also
observed.

Cognitive neuroscientists offer us findings from studies based on a new assumption: “The mapping
between physical activity in the brain and its functional state is such that when two experimental
conditions are associated with different patterns of neural activity, it can be assumed that they have
engaged distinct cognitive functions” (Rugg, 1997, p. 5). For example, Ravi Menon and his colleagues
at the Robarts Institute at the University of Western Ontario have been instrumental in using fMRI
techniques to allow a clear delineation of the human primary visual cortex in the occipital lobe
(Menon et al., 1993). As well, a review of 275 PET and fMRI studies (Cabeza & Nyberg, 2000)
summarizes how different areas of the brain are active for different cognitive functions such as
attention, perception, imagery, language, and memory.

The strategy used to produce a functional map of the brain is based on a concept introduced by
Dutch physiologist Franciscus Donders in 1868/1969. He suggested a general method for measuring
a cognitive process. This logic has been adapted to develop functional maps of cognitive processes in
the brain. He measured the time it took a person to respond (by making a keypress) to a light and
subtracted this from the time needed to respond to a particular colour of light. His experiment
revealed that discriminating colour requires about 50 msec of cognitive processing. In neuroimaging
studies the same logic, termed subtraction technique, has been applied to isolate the brain region(s)
contributing to a given cognitive process. That is, the relative amount of activation in a particular
brain region needed for a given cognitive task can be measured by subtracting a control state
(responding to a light) from a task state (discriminating colour). Figure 2-11 shows examples of how
this logic has been applied to isolate brain region contributing to various cognitive functions.

In 2000, Daniel Bub at the University of Victoria wrote a review of subtractive logic and other issues
relating to fMRI, and proposed other methods of data analysis. Borowsky and colleagues (2005) at
the University of Saskatchewan have applied a different statistical approach than subtractive logic to
understanding how the brain processes information. As well, researchers at the Rotman Research
Institute in Toronto have pioneered a new approach to understanding brain activation that does not
involve subtraction logic; rather than focus on single regions of the brain active for a given cognitive
task, a network analysis of brain region whose activity co-varies across different tasks is conducted
using multivariate statistics (McIntosh et al., 1996). Using these new methods of analysis of fMRI
data, researchers are able to paint a clearer picture of how the human brain achieves cognitive
processing—and capture just how dynamic the human brain really is, coordinating functions across
different brain regions depending on the given environment, task constraints, and motivations at the
time of processing information.

Brain imaging and recording techniques certainly include a lot of acronyms! How can the novice
keep them all straight? One way to do so is to categorize the techniques according to the kind of
information they provide. CAT and MRI scans yield neuroanatomical information. PET, SPECT, and
fMRI provide dynamic information about how blood flows during various cognitive activities. MEG,
EEG, and ERPs all measure electrical activity during cognitive activities. In the chapters to come, we
will see examples of studies that have made use of each of these techniques to investigate the
neural underpinnings of different cognitive activities.

What is important to remember is that cognitive processes are implemented in human brains. Some
researchers make an analogy between human minds and computers; in this view, the brain is the
“hardware” (“wetware”) and the cognitive processes the software. Although the two aspects of
functioning can be distinguished, to really understand either we must have some familiarity with
both, and with how they interact.