Murderous Minds (13 page)

Hot Scans

Before fMRI scanning was developed in the 1990s, the brains of violent volunteers were imaged using a more invasive procedure: positron emission tomography or PET scanning. Unlike fMRI, PET scans are considered invasive because they require an intravenous injection of a radioactively labeled tracer just before the scan begins. When the radioactive tracer is attached to glucose, the brain’s fuel, it travels to the brain and is preferentially taken up by active nerve cells. Less active cells don’t require as much energy. As the radioactive material decays, it emits subatomic particles called positrons. When positrons are emitted, radiation in the form of gamma rays is released. The scanner detects these rays. A computer analyzes the pattern of gamma ray release and translates it into an image of the brain. The more radiation that is detected in a particular region, the brighter that region is in the image.

The radioactive tracer is short-lived. It quickly becomes nonradioactive and poses no significant health threat. But few folks are thrilled by the idea of receiving a radioactive IV, no matter how harmless it is. Also, it is more trouble for researchers to deal with short-lived radioisotopes on a routine basis than it is to get people ready for an fMRI scan. In the past, PET scanners produced images less sharp than modern MRI machines. That has changed as PET scanning technology has improved. The spatial resolution of brain scanners depends on the quality of the machine, imaging time, calibration, and other technical factors; but a regular PET scanner can now have twice the resolution of a typical fMRI machine.
¹⁷

The advantages of PET include the fact that any molecule that can be radioactively tagged and introduced safely into the brain can be imaged. The resulting image reflects a straightforward physiological mechanism. Where a labeled molecule—whether a neurotransmitter, sugar, or drug—ends up in the brain is where you find it emitting its radioactive signal.

A functional MRI machine extends MRI technology almost like a video camera extends the capability of a still camera; like PET, it provides images of the brain in action. Despite what tabloid news sources claim, however, it cannot reveal what a brain is thinking, reasoning, or feeling. It cannot identify specific brain sites responsible for emotions, beliefs, or strong feelings such as love, hate, fear, anger, conservatism, liberalism, or other emotions. Those are interpretations applied by observers. The brain is far too dynamic, and its different subdivisions too interactive, to reduce higher cognitive functions or philosophical outlooks to individual central nervous system addresses. But the technology can show brain activity in the form of increased blood flow to specific brain regions. This method is known as BOLD (Blood Oxygenation Level Dependent) contrast imaging.

More-active brain cells use more energy, require more oxygen, and produce more byproducts, than less-active brain cells. The brain consumes an impressive 20 percent of the calories you eat and 20 percent of the oxygen you breathe. When neurons in a specific region of the brain become more active, they need more oxygen and more energy in the form of glucose.

Blood flow increases to meet these needs. Used blood has delivered its oxygen to nerve cells and is heading back to the heart and lungs to be refreshed, to pick up more oxygen. fMRI takes advantage of the fact that the molecule in red blood cells that carries oxygen, hemoglobin, has different effects on the magnetic resonance signal coming from hydrogen in nearby water molecules, depending on whether or not it is bound to oxygen. When hemoglobin has lost its oxygen and become deoxyhemoglobin, it dampens the signal. Hemoglobin that is carrying oxygen, oxyhemoglobin, does not dampen the signal.

When fresh blood transporting lots of oxyhemoglobin flows into a hard-working brain region, it becomes the major type of hemoglobin in that part of the brain. The presence of lots of oxyhemoglobin decreases the signal-dampening effect of deoxyhemoglobin. Decreasing a dampening effect is like decreasing pressure on a brake: when the brake is released by an influx of oxyhemoglobin into an active brain region, the fMRI picks up a greater signal from that region compared to other regions.

Researchers exploit the differing effects of fresh and used blood on the signal picked up by the MRI machine. But does the resulting image reflect
nerve-cell activity merely responding to a direct stimulus, or does it reflect nerve-cell activity processing information? Neurons responding to something send nerve signals down long axons. Neurons processing information provided by stimuli generate a lot of local electrical signaling activity.

In 2001, scientists at the Max Planck Institute for Biological Cybernetics, Tübingen, Germany managed to get a more precise idea of what type of nerve-cell activity fMRI detects. They did it by simultaneously measuring neuronal electrical activity in monkeys undergoing brain scans. They found that fMRI detects increased activity due to neurons processing information rather than responding to it.
¹⁸
BOLD signals detected by fMRI, later research showed, indeed seem to be more closely associated with activity taking place in and around synapses than in spikes generated along long axons.
¹⁹

This ongoing process which reflects the flow of blood to specific spots in the brain adds a dimension to the 3-D image produced by the MRI machine. The additional, fourth dimension allows the fMRI machine to produce images that show the relative metabolic activity of brain structures in time (a little bit delayed) as well as in space.

Researchers using fMRI create artificial color scales and assign different colors to brain regions depending on how much blood flows to them. The artificially colored results in fMRI images highlight parts of the brain that show more or less activity when performing a mental task or responding to a stimulus. The bright reds and yellows seen on fMRI images in news stories obviously do not depict what you would see if you could peer inside someone’s skull to directly observe the brain surface or subsurface structures. It might be less flashy and even more accurate to Photoshop numbers indicating the amount of blood flow to brain regions of interest, but that would be more challenging to interpret. The technology’s usefulness and popularity stems from its ability to provide images of brain anatomy coupled with function. The reliability, significance, and relevance of the resulting pictures are determined by the scientists who design and carry out the experiments and then subsequently by anyone who views and interprets the results.

A 2010 report by Carla Harenski, Ph.D., and her co-authors at the MIND Research Network in Albuquerque, N.M. illustrates a representative study made possible by fMRI technology.
²⁰
Before considering this report, recall the last time you saw a picture in the news of someone attacking another person. If you were asked to decide how good or bad, right or wrong, justified or unjustified the assault was, then brain cells in some parts of your brain involved in making moral decisions (see Figure 7 for candidates) presumably increased their activity as you made your decision.

Harenski and colleagues scanned the brains of sixteen criminals with low psychopathic traits as they considered images with moral connotations. Examples of some images used to evoke moral judgments include depictions of a physical assault in which one person attacks another, or a crime such as a home break-in. The results showed that a part of the brain called the amygdala appeared to “light up” (that is, experienced increased blood flow) when these men looked at pictures that most people agree have moral implications. The same thing happened when twenty-eight non-criminals with no psychopathic traits evaluated the pictures.

When another group of sixteen incarcerated men with high levels of psychopathic traits considered the same pictures, however, the resulting images of their brain function differed considerably from the other two groups. A glance at Figure 13 indicates that criminals with high psychopathic traits appear to have decreased levels of activation in their amygdalae when looking at these images of moral violations.

The presence of psychopathic traits seems to affect the amygdala’s response to other select stimuli in a similar way, even in the presence of a major psychiatric disorder such as schizophrenia. An fMRI study carried out in Australia, for example, looked at the effect of psychopathic traits on amygdala response to images of fearful faces. The researchers concluded that violent patients with schizophrenia and high psychopathy scores showed decreased amygdala responses compared to violent patients with schizophrenia and low psychopathy scores.
²¹

It is important to remember that such changes detected in fMRI studies, although real, may sometimes represent small differences between groups. Also, individual variations in human brain activity may vary greatly from one person to another, and in one person from day to day, or even from hour to hour. Consequently, fMRI results are often shown as averaged changes displayed on a single representative brain image.

Limitations

An activated brain region—one involved in recognizing a potential threat raised by a disturbing photograph, for instance—may be marked by less than a five percent increase in blood flow compared to uninvolved brain regions. As mentioned, the brain uses a lot of energy even when it seems to be resting. This means it is a highly active organ metabolically. When it “goes to work” on a particular task, the increased flow of blood delivering oxygen and glucose to a particular area is not as large as you might predict. It is not easy to pick out small changes in such a complex and active structure, because the brain’s background activity is like “noise” that can make it a challenge to pick out the particular signal you might be interested in. It is, after all, a system capable, in the best instances, of creating and appreciating great literature, art, and scientific insights, and, in the worst, of concocting and executing schemes to gain control, victimize, and even terrorize individuals or entire populations.

Craig M. Bennett, Ph.D., is a cognitive neuroscientist at the University of California, Santa Barbara with an interest in magnetic resonance imaging methods and a talent for explaining the complexity of his field to scientists and nonscientists alike. He and his colleagues, for example, used a memorable approach to emphasize the precautions and steps researchers must take to make sure they end up with valid brain images using fMRI. They produced an unusual abstract presentation entitled “Neural Correlates of Interspecies Perspective Taking In the Post-Mortem Atlantic Salmon: An Argument for Multiple Comparisons Correction.”
²²
The poster presentation included authentic fMRI images of a dead fish whose brain clearly shows activity in the form of a dab of overlaid red color.

The apparent mental activity coincided with the fish’s participation in a brain-scanning experiment in which the subject was presented with a series of photographs of human faces and asked to identify the emotion expressed in each picture. To paraphrase John Cleese in
Monty Python’s Flying Circus
’s famous Dead Parrot Sketch,
²³
the piscine subject of this demonstration—it bears repeating—was no more, had ceased to be, expired, gone to meet its maker, was bereft of life, its metabolic process now history. If it had not been slid into an fMRI machine, it would have been pushing up the daisies or else been prepared for dinner.

This tongue-in-cheek experiment, of course, was conducted for a serious

purpose. It was performed to emphasize to users of the increasingly popular fMRI technology that extreme care must be taken to avoid false positive results—that is, indications that something is happening when it is not. “Images of brain activity only have meaning when acquired using the correct experimental design and interpreted using the correct analyses,” the authors of the one-time experiment noted.

“Not only are the steps not standardized, they are easily manipulated by a person with knowledge of the technology. Color coding, for example, can be arbitrary and may present the illusion of huge differences in some aspect of brain activity, when little actually exists,” the authors warned back in 2009. But things have improved since then. “If I give you a dart and have you throw it at a dartboard [without aiming], there is a certain probability (let’s say 1%) that you will get a bull’s-eye,” Bennett explained four years later.
²⁴
“If I give you 30,000 darts, then you are going to get approximately 300 bull’s-eye hits just because of random chance. The same is generally true in neuroimaging. We are testing for significant results across 30,000+ voxels [the 3-dimensional MRI equivalent of a flat screen’s pixels], and some will be significant by chance. To really know the true probability of a false positive in our data, we need to do a statistical correction for this multiple testing problem. In 2008, 30–40% of papers did not utilize any form of correction. In 2010, it was less than 10%.”

The field of psychopathy research using fMRI imaging is young, but it has made enough progress that its findings are clearly fitting into a pattern. They are being reproduced in different labs and so are laying the groundwork for one day developing a good understanding of how the brains of criminal psychopaths differ from the brains of non-psychopaths. As we will see, the evidence points to problems involving many parts of the brain that lie below its surface and with regions of the cerebral cortex directly connected to them. However, as always, consumers need to be aware of the limitations of the techniques used to gain these insights. In the future, the progress and the limitations of this research are likely to produce some interesting challenges for a society concerned about the personal cost of living with psychopaths—both criminal and those who never commit a crime.