American Journal of Law & Medicine

The brain and behavior: limitations in the legal use of functional magnetic resonance imaging.(Brain Imaging and the Law)


Brain imaging is one of the most remarkable technological advances towards understanding the relationship of behavior to brain anatomy and physiology. Brain images provide insight to understanding behavior. Additionally, the images themselves carry great impact, particularly when used to show differences in either the anatomy or the biological functioning of two different brains. For these reasons, brain images have increasingly been used in both criminal and civil trims.

After describing some general features of brain imaging, we will focus on functional magnetic imaging (fMRI), as many believe this technology has the most potential for advancing our understanding of how parts of the brain function, including perhaps linking specific functions with cognition and behavior. Brain imaging as a field is vast and therefore our discussion will be limited. First, we will assess the advantages and limitations of fMRI, including research efforts towards standardizing equipment thereby assuring reliability and reproducibility. Second, we will address the extent to which fMRI links brain function to behaviors. We will then discuss the enhancing benefits of combining of fMRI with other imaging technologies, like electroencephalography (EEG) (1) and magnetoencephalography (MEG). (2) Finally, we will consider the relevance of fMRI for law, especially the potential for assessing personal responsibility and detecting truth and lies during interrogation.


Whether structural or functional, brain imaging can involve a variety of methods. Each method is geared towards detecting and measuring specific signals of some property related to the brain with a detection device that is usually outside of the brain. (3) Signals that are detected vary from electrical activity in the brain, (4) to magnetic dipoles induced from outside the brain, (5) to radioactive events through the injection of radioactive tracers. (6) These signals measure different properties of brain tissue and vary widely in parameters, including specificity, sensitivity, temporal and spatial resolution, and in their fidelity towards reflecting the physiological processing being studied. (7) In addition, the images created are not immutable. Although some properties of the brain are largely invariant at the functional level, adaptation in the service of homeostasis and as a necessary response to the varying environment will insure that the micro-anatomy must be ever-changing.

Regardless of the source, signals result in distinctive patterns that would seem to allow for comparison among individuals. For example, an EEG recording of a particular area of the brain of a subject shows a repetition of brain activity, which is then recorded on a graph. Thereafter, the record can be compared with the electrical activity of another record comprised of composite recordings of many 'normal' brains. (8) These signals can, as in the case of positron emission tomography (PET), single-photon emission computed tomography (SPECT), and structural and functional MRI, be fed into a computer that stores the data and uses the information to form or reconstruct an image of the brain. (9) In order to enhance visual impact, the image produced is generally color-coded along a spectrum from blue and green to red and yellow to reflect the varying degrees of activity in specific regions of the brain. (10)

Functional brain imaging may have an important role in our understanding of the relationship of brain processes to behavior. (11) It offers the possibility to couple an image with neurocognitive and emotional functioning. (12) As we will see, some of these functioning imaging modalities, particularly PET and fMRI, go further to integrate brain function with features of molecular and neural circuitry information. (13) Basic functional imaging can be achieved through a variety of methods, from PET and SPECT to quantitative EEG, MEG, and fMRI, each of which measures a different physical property of the brain. To date, however, among these different imaging technologies, PET and fMRI have been the most technologically advanced, and have had the broadest application in the courtroom. (14)

In the case of PET, the most frequently studied biological process has been energy metabolism. (15) This is primarily because energy metabolism is closely linked to brain function, although in a very complex way. (16) Energy metabolism and, therefore, brain function, is revealed through the study of three components of energy, which are normally physiologically coupled. (17) These components are glucose metabolism, oxygen metabolism, and cerebral blood flow. (18) Glucose metabolism is studied through the use of an analogue of glucose (i.e. deoxyglucose) labeled with a radiotracer such as Flourine-18 or Carbon-11. (19) Oxygen metabolism is investigated with the use of Oxygen-15, and cerebral blood flow with Oxygen-15 labeled water. (20) Because it is a tracer method, PET has the distinct advantage of being thus far the best modality for the detection of a wide variety of biochemical processes. (21) In fact, its only limitation is chemical ingenuity and its inherent high sensitivity. (22) Furthermore, one of its advantages is that PET has a high degree of quantification accuracy regarding changes pre- and post- intervention in brain regions with altered brain perfusion or metabolism. (23) Unfortunately, interpretable PET data are almost never available for any individual prior to the incident, behavior or brain insult that led to the legal proceeding. (24) Nevertheless, in current standardized settings, which are rigorously defined, PET data are very reproducible. (25)

MRI, on the other hand, relies upon a particular physical property of the hydrogen atom--the inherent spin of its atomic nucleus--that enables all of the hydrogen nuclei in water to be oriented similarly in the presence of a very strong homogeneous magnetic field. When the magnetic field is turned off, these nuclei will "relax" in a manner that is characteristic of their chemical and physical environment. In the body, water is by far the most common molecular entity containing hydrogen and hence, when we look at a typical MRI presentation we are usually looking at the differing environments of water.

fMRI operates under the principle that changes in the brain's hemodynamics, which relate to mental operations, can be detected and mapped using basic MRI instrumentation. (26) At this time, the most widely used method to measure cerebral blood flow using MRI has been the Blood Oxygen Level Dependent (BOLD) technique. (27) This technique depends on the 70-year-old observation that the properties of hemoglobin in a strong magnetic field are dependent upon its state of oxygen saturation. (28) The underlying physiological notion is that increased neural activity in a particular brain region results in more consumption of oxygen from the blood near these neurons. (29) Accompanying the increased oxygen consumption are increases in blood flow and blood volume of the local vasculature of the activated regions of the brain. (30) The consequence is that the "blood near a region of local neuronal activity ... has a higher concentration of oxygenated hemoglobin than blood in locally inactive regions." (31) But recent research indicates that the sequence is open to some debate. (32) At least partly, the belief that the augmentation of cerebral blood flow (CBF) is in response to the energy requirements of the brain for nutrients and oxygen to maintain functional activation may still hold. (33) But some research has also shown that glycolytic demands for energy metabolism may not be the primary influence of blood flow. (34) Rather, the evidence is leaning in the direction of relating blood flow to the need for removing the toxic products from metabolism, such as the accumulation of lactate. (35) By the same token, and perhaps not inconsistent with this, there is some evidence that the hemodynamic response may be related to the release of presynaptic neurotransmitters, thereby reflecting local signaling. (36)

fMRI is similar to PET in that it accurately localizes signal sources, thereby more closely identifying regions of the brain in terms of anatomy and function. (37) Its most important application to date has been to map the hemodynamic responses to defined cognitive and affective stimuli to determine the anatomical loci subserving specific brain functions in the cognitive, behavioral, and affective domains. (38) The grossly oversimplified underlying assumption has been that cognitive functions are basically located in focal brain regions, (39) though in fact that is unlikely the whole picture. Evidence from brain studies points to the notion that most complicated behavioral and psychological processes are not located in a single brain center. (40) Neuronal circuitry regarding any one cognitive operation most likely extends into more than one circuitry, though in fact the concept of "localization" may refer to functions that are causally connected to specific neuronal circuits. (41)

It is simplistic to consider the concept of localization without a consideration of the "inverse problem." Simply stated, the inverse problem is from where and when in the brain did the measured signal emanate? All of the imaging techniques in use today are measuring signals that emanate from within the brain with detectors that are outside of the brain. Some of these signals are largely independent from their distance to the detectors, while other signals are exquisitely sensitive to their sites of localization in the brain (e.g. superficial versus deep). Some of these signals emanate from a very discrete source while others are widely distributed. The sensitivity and specificity of the measurements both independent of and in relation to the physiological process being observed vary enormously from one technique to another. It is not only the fidelity and the spatial resolution that must be taken into account, but the temporal resolution as well. Electrophysiological events can be recorded in milliseconds, (42) while glucose utilization is measured over many minutes (43) and fMRI is intermediate. (44) In this sense of spatial localization, both modalities of PET and fMRI are superior to imaging methods relying on electrophysiological properties of the brain, such as EEG and MEG. EEG and MEG are capable of relating neural events to cognitive sequences that lead to the execution of a task, (45) and they both have relatively high temporal resolution, which is an advantage over fMRI and PET. (46) In the case of MEG, however, the detection seems mostly of small cell clusters located on the surface of the brain. And, although EEG can detect changes in neuronal activities that occur in milliseconds, neither EEG nor MEG is able to obtain three-dimensional spatial patterns of neural activity. In contrast, fMRI provides very good spatial resolution, though at times inadequate temporal resolution. (47)


Generally, fMRI is more readily available to researchers and clinicians and provides greater detail and resolution than comparable imaging methods. Additionally, fMRI has two distinct advantages over its closest competitor, PET, for imaging brain activity directed to specific sensory responses or cognitive tasks. (48)

First, the production of the signal for measurement does not require the introduction of radioactive isotopes. Therefore, it can be repeated often, if necessary, on the same individual without the concern of adverse radiological effects. (49) In that sense fMRI is essentially non-invasive; in its most frequent usage it relies on the presence of a by-product of neural activity, which is oxygen utilization. (50) This is the BOLD method, which depends on the variation in the content of deoxyhemoglobin of the cerebral vessels at or near the site of neuronal activity, which is oxygen utilization. (51) The BOLD contrast mechanism is a result of magnetic field inhomogeneities brought about by deoxyhemoglobin. (52) Relative to brain tissue, deoxyhemoglobin is slightly paramagnetic; oxyhemoglobin is isomagnetic, so that blood vessels that have primarily oxygenated arterial blood essentially have little effect on the magnetic field in surrounding tissue. (53) In contrast, blood that is partially deoxygenated in capillaries and veins has a distorting effect on the magnetic field. …

Log in to your account to read this article – and millions more.