Advances in MRI and computing technologies led from anatomical to functional brain-imaging techniques, which allow investigators to measure the amount of blood, oxygen, and glucose the brain uses as subjects or participants solve cognitive problems. When a brain region is active, the amount of blood, oxygen, and glucose flowing to the region increases. It therefore is possible to infer changes in brain activity by measuring either blood flow or levels of the blood’s constituents, such as oxygen, glucose, and iron. Three techniques developed from this logic are functional MRI, optical tomography, and positron emission tomography.

230

As neurons become active, they use more oxygen, resulting in a temporary dip in the blood oxygen level. At the same time, active neurons signal blood vessels to dilate to increase blood flow and bring more oxygen to the area. Peter Fox and colleagues (1986) discovered that when human brain activity increases, the extra oxygen produced by increased blood flow actually exceeds the tissue’s needs. As a result, the amount of oxygen in an activated brain area increases.

Oxygen is carried on the hemoglobin molecule in red blood cells. Changes in the ratio of oxygen-rich hemoglobin to oxygen-poor hemoglobin alters the blood’s magnetic properties, because oxygen-rich hemoglobin is less magnetic than oxygen-poor hemoglobin. In 1990, Segi Ogawa and his colleagues showed that MRI could accurately match these changes in magnetic properties to specific brain locations (Ogawa et al., 1990). This process, functional magnetic resonance imaging (fMRI), signals which areas are displaying changes in activity.

Figure 7-17 shows changes in the fMRI signal in the visual cortex of a person who is being stimulated with light. When the light is turned on, the visual cortex (bottom of the brain images) becomes more active than during baseline (no light). In other words, functional changes in the brain are inferred from increases and decreases in the MRI signal produced by changes in oxygen levels.

When superimposed on MRI-produced anatomical brain images, fMRI changes in activity can be attributed to particular structures. The dense blood vessel supply to the cerebral cortex allows for a spatial resolution of fMRI on the order of 1 millimeter, affording good spatial resolution of the brain activity’s source. On the other hand, because changes in blood flow take as long as a third of a second, the temporal resolution of fMRI is not as precise as that obtained with EEG recordings and ERPs.

fMRI also has the disadvantage that subjects must lie motionless in a long, noisy tube, an experience that can prove claustrophobic. The confined space and lack of mobility also restrict the types of behavioral experiments that can be performed. Nonetheless, fMRI is a major tool in cognitive neuroscience.

231

The living brain is always active, and researchers have succeeded in inferring brain function and connectivity by studying fMRI signals when participants are resting, that is, not engaged in any specific task. This signal, resting-state fMRI (rs-fMRI), is collected when participants are asked to look at a fixation cross and to keep their eyes open.

The scanner collects brain activity, typically for at least 4-minute blocks. Researchers are attempting to shorten the period by increasing the strength of the static magnetic field and developing more sensitive coils. Statistical analysis of the data entails correlating activity in different brain regions over time. Although rs-fMRI is still in its growth phase, investigators already have identified many consistent networks of brain activity and abnormalities in disease states such as dementia and schizophrenia (Van den Heuvel & Hulshoff Pol, 2010).

Research Focus 7-1, Tuning into Language, describes a brain-imaging study that used functional near-infrared spectroscopy (fNIRS) to investigate newborn infants’ responses to language. fNIRS is a form of optical tomography, a functional imaging technique that operates on the principle that an object can be reconstructed by gathering light transmitted through it. One requirement is that the object at least partially transmit light. Thus, optical tomography can image soft body tissue, such as that in the breast or the brain.

In fNIRS, reflected infrared light is used to determine blood flow because oxygen-rich hemoglobin and oxygen-poor hemoglobin differ in their absorption spectra. By measuring the blood’s light absorption it is possible to measure the brain’s average oxygen consumption. So fNIRS and fMRI measure essentially the same thing but with different tools. In fNIRS, an array of optical transmitter and receiver pairs are fitted across the scalp, as illustrated in Figure 7-18A.

The obvious advantage of fNIRS is that it is relatively easy to hook subjects up repeatedly and record from them for short periods, from infancy to senescence. The disadvantage is that the light does not penetrate far into the brain, so researchers are restricted to measuring cortical activity (Figure 7-18B). The spatial resolution is also not as good as with other noninvasive methods, although NIRS equipment now uses over 100 light detectors on the scalp, which allows acceptable spatial resolution in the image. NIRS has been used to differentiate cancerous from noncancerous brain tissue. This advance should lead to safe, extensive surgical removal of brain cancers and improved outcomes (Kut et al., 2015).

232

Researchers use positron emission tomography (PET) to study the metabolic activity of brain cells engaged in processing brain functions such as language. PET imaging detects changes in the brain’s blood flow by measuring changes in the uptake of compounds such as oxygen and glucose (Posner & Raichle, 1997). A PET camera, like the one shown in Figure 7-19, is a doughnut-shaped array of radiation detectors that encircles a person’s head. A small amount of water, labeled with radioactive molecules, is injected into the bloodstream. The person injected with these molecules is in no danger because those used, including the radioactive isotope oxygen-15 (¹⁵O), are very unstable. They break down in just a few minutes and are quickly eliminated from the body. (Most of the oxygen in air we breathe is the stable ¹⁶O molecule.)

Radioactive ¹⁵O molecules release tiny, positively charged subatomic particles known as positrons (electrons with a positive charge). Positrons are emitted from an unstable atom because it is deficient in neutrons. The positrons are attracted to the negatively charged electrons in the brain, and the collision of the two particles leads to annihilation of both, which produces energy.

This energy, in the form of two photons (a unit of light energy), leaves the head at the speed of light and is detected by the PET camera. The photons leave the head in exactly opposite directions from the site of positron–electron annihilation, so annihilation photon detectors can detect their source, as illustrated in Figure 7-19. A computer identifies the coincident photons and locates the annihilation source to generate the PET image.

The PET system enables blood flow measurement in the brain because the unstable radioactive molecules accumulate there in direct proportion to the rate of local blood flow. Local blood flow in turn is related to neural activity because potassium ions released from stimulated neurons dilate adjacent blood vessels. The more the blood flow, the higher the radiation counts recorded by the PET camera.

Sophisticated computer imaging can map blood flow in the brain when a person is at rest with closed eyes (Figure 7-20). The resting-state map shows, in a series of frames, where blood flow is highest. Even though the distribution of blood is not uniform, it is still difficult to conclude much from such a map because the entire brain is receiving oxygen and glucose.

233

But PET researchers who are studying the link between blood flow and mental activity use a subtraction procedure. They subtract the blood flow pattern when the brain is in a carefully selected control state from the pattern of blood flow imaged when the subject is engaged in the task under study, as illustrated in the top row of Figure 7-21. This subtraction process images the change in blood flow between the two states. The change can be averaged across subjects (middle row) to yield a representative average image difference that reveals which brain areas are selectively active during the task (bottom). PET does not measure local neural activity directly; rather, it infers activity on the assumption that blood flow increases where neuron activity increases.

A significant limitation of PET is that radiochemicals, including so-called radiopharmaceuticals used in diagnosing human patients, must be prepared in a cyclotron quite close to the scanner because their half-lives are so short that transportation time is a severely limiting factor. Generating these materials is very expensive. But in spite of the expense, PET has important advantages over other imaging methods:

Functional Brain Imaging

Before you continue, check your understanding.

Question 1

Question 2

Question 3

Question 4

Answers appear in the Self Test section of the book.