Brain Images as Legal Evidence

This paper explores whether brain images may be admitted as evidence in criminal trials under Federal Rule of Evidence 403, which weighs probative value against the danger of being prejudicial, confusing, or misleading to fact finders. The paper summarizes and evaluates recent empirical research relevant to these issues. We argue that currently the probative value of neuroimages for criminal responsibility is minimal, and there is some evidence of their potential to be prejudicial or misleading. We also propose experiments that will directly assess how jurors are influenced by brain images. Brain images are becoming more and more common in courts. Feigenson (2006) found 130 reported opinions involving PET and/or SPECT evidence but only 2 reported opinions citing fMRI evidence. Helen Mayberg, however, has served as an expert witness in over 50 trials in recent years, many of them involving fMRI evidence, and a number of judges have informally told us that evidence from neuroscience, including fMRI, has become standard in capital sentencing. Some lawyers and neuroscientists are critical of this trend. A few have even suggested in conversation a temporary moratorium on brain images as legal evidence in criminal trials, except possibly in capital sentencing. This paper will explore the prospects for some uses of data from brain imaging in the courts. Whether brain images should be admitted into trials depends, of course, on how probative they are for specific legal issues and on whether they are likely to mislead fact-finders in trials. We will address these topics in turn after illustrating the variety of uses of brain images in law. 1. W H A T C O U L D B R A I N I M A G E S B E L E G A L E V I D E N C E F O R ? Brain images could conceivably be used for many different purposes within the legal system. Neuroscientific studies involving structural or functional brain images DOI: 10.3366/E1742360008000452 E P I S T E M E 2008 359 Walter Sinnott-Armstrong et al. might, for example, be used to argue for or against certain legislation or prison policies. They might also be used to justify predictions of misbehavior in parole hearings. We will, however, focus on the use of functional brain images in courts. One proposed use of functional brain images in trials is for mind-reading. When brain scans are used to detect lies or deception, for example, they are supposed to detect whether a person has a mental state of belief in what they say. A few companies (Cephos and NoLieMRI) already offer methods of lie detection using fMRI. Although their results are uncompelling to date, EEG data were admitted as evidence against lying in 2001 by Iowa District Court Judge Tim O’Grady in the case of Terry Harrington. Brain images might also be introduced as evidence of mental states other than deception, such as bias in jurors, consciousness in cases of end of life issues, and pain and suffering in tort plaintiffs or applicants for disability benefits. Another possibility is to introduce brain images as evidence not of temporary mental states but of more stable mental traits or capacities. Mental capacities might be relevant to competence to stand trial or to be executed or they might be relevant to criminal responsibility if they involve incapacities to gain requisite knowledge or to form requisite intentions that are part of the mens rea of most crimes. Neuroscience might also become relevant to whether adolescents and people with brain damage lack substantial capacity to conform their conduct to the law or whether psychopaths, for example, lack substantial capacity to appreciate wrongfulness. Evidence of such capacities might be relevant to an insanity defense, in the guilt phase, or in the sentencing phase of a criminal trial. This list of potential uses is incomplete, but it gives some sense of the wide range of possible uses of neuroscientific evidence in legal trials. Many of these uses are speculative, and they need not all be treated alike. Each proposed use of neuroscientific evidence needs to be assessed carefully in context and on its own. For the sake of simplicity, we will focus here on functional brain images used by the defense in a criminal trial to reduce responsibility. These uses might occur in the guilt phase to challenge an element of the crime charged, in an insanity defense, or in the sentencing phase to argue for a lighter sentence. 2. L E G A L S TA N DA R D S O F E V I D E N C E Whatever legal issue is at stake, to introduce a brain image as evidence in a legal trial, either side needs to meet standards for demonstrative evidence (such as exhibits) as well as standards for scientific expert testimony. The testimony is needed in order to interpret the images during the trial. Although it would also be important to ask when neuroscientists (especially cognitive neuroscientists) meet requirements for expert testimony, we will restrict our discussion here to whether or when functional brain images may be admitted under the rules governing demonstrative evidence. 360 E P I S T E M E 2008 BRAIN IMAGES AS LEGAL EVIDENCE Most courts follow something like the Federal Rules of Evidence: FRE 401: “relevant evidence” means evidence having any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence. FRE 403: Although relevant, evidence may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence. To apply FRE 403 to brain images in criminal trials, courts must answer three central questions: (1) How probative for criminal responsibility is the brain image? (2) How dangerous (that is, prejudicial, confusing, misleading, or needless) is the brain image? (3) Does its danger substantially outweigh its probative value? To answer these questions, we need to understand, first, how brain images are constructed. Only then can we determine their probative value and whether they confuse or mislead. 3. W H A T I S A B R A I N I M A G E ? There are many types of brain images, but usually the term refers to images derived from noninvasive techniques for measuring structural or functional properties of the brain. A number of such techniques exist, including PET, SPECT, MEG, DTI, structural MRI, and functional MRI (fMRI). We will focus here on fMRI, though our main points will apply as well to other functional brain imaging techniques. The most commonly used fMRI techniques measure changes in the ratio of oxygenated to deoxygenated blood (the BOLD signal). This signal is closely related to blood flow and bears a complicated relation to neural activity. These relations are well-documented although not yet completely understood. Inferences about brain activity are typically made by designing experiments that contrast the MR signal measured during two different tasks. Ideally, the tasks differ in one respect, and the location and magnitude of the difference in measured signal is attributed to brain activity involved in the difference in task performance. For instance, one task might involve processes A-E, and another may involve processes A-D but not E. The difference in signal is thus interpreted to be involved in process E. In practice, there are almost always a number of differences among the tasks. With enough psychological sophistication, these can be modeled, although they are not always easily assessed. There are also many minor differences across trials while performing the same task, such as differences in processing individual stimuli; and the signal itself is noisy. When enough stimuli are presented, these minor differences will wash out in the statistics. The difference in MR signal between task conditions is usually quite small, often less than 1%, but with enough data even such small differences can be statistically significant. E P I S T E M E 2008 361 Walter Sinnott-Armstrong et al. The discovered difference in MR signal is often presented as a brain image. These images are usually constructed by superimposing colored pixels on a grey-scale picture of a standard brain in order to indicate where signal was higher (usually red or yellow) or lower (usually blue) than in a contrast state. The resulting fMRI images look something like photographs, but they are not photographs. Instead, they are constructions from highly abstract numerical data about magnetic properties. Brain activity and blood flow are not brightly colored, and the brain does not really “light up” when active. It is important to bear in mind that brain images are simply a vivid way to represent the location and magnitude of statistical differences in signal across large data sets. (For more detail, see Roskies 2007, 2008.) 4. A R E B R A I N S C A N S P RO BAT I V E O F C R I M I N A L R E S P O N S I B I L I T Y ? The Federal Rules of Evidence do not explicitly define probative value. However, one standard, respected textbook defines probative value as degree of relevance: Remember that evidence is “relevant” if it has “any” tendency to make the fact of consequence more or less probable; probative value measures the strength of the effect on the probabilities, even if only in general terms like “highly,” “somewhat,” or “minimally” probative. (Allen et al. 2006, 135) This account makes probative value equivalent to a relative conditional probability. We doubt that the issue is this simple, because values enter into the equation in ways that we will see. Still, a good starting point for assessing the probative value of any evidence is to ask how much the evidence increases the probability of some fact that matters. To apply this standard to brain images, we need to consider the precise nature of the information that is presented in the image and also which fact the image is supposed to be evidence for. This is an immensely complex topic. Here we can only run through five main problems that arise when trying to use brain images as evidence of facts that are relevant to criminal responsibility.