fMRI Lie Detection: Is there Science Behind Deception?

Author:  Karuna Meda
Institution:  University of Pennsylvania

Lies, truth and science

Deception is a fascinating, yet unsettling, trait that pervades all cultures and societies, festering in the fissures of fragile human relations. Lies can take various forms, whether it is the blatant manipulation of facts or subtle omissions; in either form, deception is the concerted suppression of the truth. The search for the truth-or the lack thereof- is something social scientists have pursued for centuries, in an effort to relieve the distrust and deceit that destabilize social relations. However, it is difficult to identify a liar with consistency and certainty. What behavioral markers does one look for? Shifty eyes, a shaky voice, and trembling hands are not sufficient evidence in the context of a courtroom prosecution. It is true that the anxiety of deceit is channeled in these peripheral physical actions, but it was necessary for those interested in detecting deception to approach the highly subjective issue of deception in a more objective fashion. Consequently, these efforts turned to the purely objective approach - expert' science. Thus the following seminal questions were posed- is there science behind lying? Are there certain biological processes directly involved in deception? More importantly, is there a way to detect these physiological parameters? Numerous scientists and technologies have been utilized by social scientists and lawmakers to answer these questions.

The world's first "physiological" lie detector was the polygraph. The machine was the mainstay of forensic lie detection for almost a century before scientists began looking at the true purported source of lying- the brain. The introduction of electroencephalography (EEG) provided great insight into the potential advantages of using neurophysiologic measures to observe the process of deception in the brain. The latest "neurotechnology" that has hit the lie detection scene is fMRI. Scientists claim that fMRI- or functional magnetic resonance imaging- can detect increased blood flow to key areas of the brain that are supposedly implicated in the "thought process of lying". This forces the quandary- how far can one equate neurobiology and emotional thought processes involved in deceptive behavior? Furthermore, at what ethical expense are we willing to use this new technology to invade the brains of the potential liars? The following discourse explores these issues while evaluating whether fMRI is the ultimate lie detector that the world has been waiting for.

Lie Detection- A coveted technology

Lie detection has become somewhat of a cultural icon; TV shows like the forensic dramas and numerous crime-drama movies attract large audiences. But what does lie detection mean in the context of global interests and crime? Without doubt, the power to detect deception would provide a perceivably safer haven for social relations.

The origins of the first lie detector are murky; scientists had conducted numerous experiments since the 18th century in various espionage efforts. However, the first modern polygraph machine can be traced back to a student in experimental psychology at Harvard University- William M. Marston- prior to 1921. In the same year, August Vollmer of California Police Department encouraged Dr. John A. Larson, a psychiatrist at Berkeley, to modify Marston's polygraph and develop a technique for continuous recording of physiological responses. Leonard Keeler, a Berkeley police officer, later patented the technology and helped publicize its use.

The instrument made the first permanent record of blood pressure, pulse, and respiration. The theory was that a person who is lying exhibits involuntary physiological responses such as increased blood pressure, perspiration, breathing, and heartbeat. Thus, the polygraph transformed our understanding of truth and lies- these concepts now had a corporeal meaning. Lawmakers were certainly convinced, and during the 1960s and 70s, the polygraph business raked in millions of dollars and was used routinely in police interrogations and as "expert witnesses"3 in court trials. Over the next decade, the technology moved into the realm of military and security agencies.

The excitement surrounding the polygraph began to fade, however, as people began to question the scientific validity of the technology. One of the major failings of the technology was the interpersonal bias between the examiner and examinee; it is very easy for the examiner to misinterpret a subject's reaction to a particular question. Furthermore, there were a number of ways one could trick' the machine, by using antiperspirants on fingertips, biting one's tongue, or any other action that can suppress a detectable physical reaction to the examiner's questions. Perhaps the most significant drawback of the polygraph was that it only recorded peripheral rather than central nervous system physiological measures of deception; that is, it only represents outward emotional manifestations of lying and not the true cognitive source- the brain. Today, polygraphs are almost never admissible in a court of law. 4

fMRI-The cartographer of cognition?

Functional magnetic resonance imaging – or fMRI - enables researchers to create maps of the brain in action as it process thoughts, sensations, memories, and motor commands. Since its introduction to experimental medicine ten years ago, fMRI has been successfully used in the prognosis of diseases like Alzheimer's and schizophrenia, and the evaluation of drug treatments. fMRIs have also been used in conjunction with neurosurgery to localize damage.5 More recently, researchers have claimed that fMRIs have given insight into the cognitive operations behind such complex and subtle behavior such as deception.

Daniel Langleben, a psychiatrist at the University of Pennsylvania, was one of the first scientists to research the technology's applications in deception. His foray into the area was, however, inadvertent. "When I was studying adolescent boys with ADHD, I was intrigued by the observation that these children had difficulty lying. Children with ADHD have impulse control problems, they cannot inhibit their reactions. As a result, the truth that is sitting there in their brains, just bubbles out". From this starting point, Dr. Langleben began to study whether there was a biological difference between truth' and lie', using fMRI. So, how exactly does fMRI project an accurate' representation of a truth' and a lie'?

In an interview, James Loughead, professor of neuropsychology at Penn and one of Langleben's collaborators, elucidated how brain fMRI is performed by placing the head or the entire body in the bore of an MRI scanner, built around a powerful electrical magnet (whose magnetic field is around 105 times stronger than the Earth's). The polarized magnetic field causes the hydrogen nuclei in water in the body to resonate and emit radiofrequency signals that are eventually reconstructed into a three-dimensional image that reflects the relative concentrations of the hydrogen nuclei in the tissues.

"The key quantification that an fMRI gives us is the haemodynamic response- or the Blood Oxygen Dependent Level (BOLD)," Dr. Loughead explained. "When local oxygen demand rises in response to increased electrical activity and metabolism, the BOLD signal changes. The BOLD signal can therefore tell us in which areas of the brain neurons are most active in response to various experiences". As an example, Dr. Loughead points to the BOLD pattern for a finger-tapping task. This simple task can be localized to the primary motor cortex consistently on a single subject basis,there is little or no room for error in this detection.


"Lying is an involuntary action; it's not something you can suppress," Loughead explained. "You need to exercise a system that is in charge of regulating and controlling your behavior when you lie more than when you say the truth". Langleben and Loughead used the Guilty Knowledge Test (GKT) in which subjects were given envelopes with a playing card in it, and they were asked to lie about the identity of the card. Once in the scanner, the subjects were shown a series of pictures of playing cards, including the one they had in their pocket, and they pressed yes' or no' to each card. When it came to the card they were hiding, they would say no'- and this would constitute a lie, as is displayed in Figures 1a and 1b.



Figure 1a. A segment from the computerized GKT adapted for event-related fMRI. Each "Truth" (2 of Hearts), "Lie" (5 of Clubs), and "Control" (10 of Spades) was presented 16 times, each Non-Target card,twice. Stimulus presentation time was 3 s, interstimulus interval,12 s, total number of presentations,88. Order of presentation was pseudorandom (randomly predetermined). (Brain Activity during Similated Deception- an Event Related Functional Magnetic Resonance Sstudy, Langleben et al, 2001)



Figure 1b: fMRI results of CIT meta-analysis (Lie > Truth). Image thresholded at voxel-height probability of p < 0.001, and cluster-probability of p < 0.05. Results are z-statistic maps displayed over MNI T1 anatomical template in radiological convention. Significant clusters of activation located in bilateral IFG, IPL, SFG, Dorsal ACC, MTG, and precuneus. (Hakun JG, Seelig D, Ruparel K, Loughead JW, Busch E, Gur RC, Langleben DD. 2008. "fMRI investigation of the cognitive structure of the Concealed Information Test." Neurocase:14(1):59-67.)

"The basis of the test is a robust human response to salience", says David Seelig, a graduate student in Langleben's lab. "If you were to present suspects with various murder weapons, including the one found at the crime scene, the reaction to salience is something that can be detected by the fMRI, and researchers can see which suspects show an increased brain response to the real weapon".

One safety precaution is that no metal objects can be worn or brought near the fMRI machine during operating hours; there have been disasters where the insuperable magnetic field of the scanner has essentially ripped out a patient's metal replacement, such as a pacemaker or any metal prosthetics.

According to the studies conducted by Langleben and Loughead, the parts of the brain that are most active during deception are the inferior frontal gyrus and the inferior parietal lobe (together they make up the prefrontal parietal network important in detecting salience) and bilateral frontal cortex. The increased blood flow to these areas, as represented by the BOLD patterns, indicated that these areas were working harder in telling a lie' than they were in telling the truth'. Langleben and Loughead have provided the desired cognitive markers that neuroscientists, psychologists, lawmakers and the general public have been looking for ever since the downfall of the polygraph. Their data also, however, raises some important questions- is deception in such an instructive manner (that is, the subject is essentially instructed to lie) really deception? As Seelig explains, the fMRI does not measure two important components of lying- the emotional response and the intention of the liar' to instill false belief in another person. "Telling a story is very different from simply responding to Yes' or No' questions; you have to maintain two realities. This requires the emotional component of deception, controlled by the limbic network." The limbic network includes the purported emotional center of the brain- the amygdala. These aspects, in addition to the cognitive component, give a more accurate representation of deception. Considering this, what are the implications of using an fMRI in legal and security milieu? If it is not a true representation, there is the risk of falsely incriminating someone, either in a situation as dire as court or as routine as job screenings.

The Thought Police

Recently, at a symposium titled "Is There Science Underlying Truth Detection?" held at the American Academy of Arts and Sciences6, Jed Rakoff, a United States district judge for the Southern District of New York, made it very clear that he will never allow an fMRI scan as evidence for deception in his court. Judges like Rakoff are worried that lie detectors may take away from the jury its role as fact-finder, specifically when assessing the credibility of a witness. One could imagine that a jury will no longer have to watch a witness testify, hunt for deceptive behavior, and decide whether or not to believe the witness, because the fMRI will isolate the "liars." What does this "mechanization," or dehumanizing of deception detection, in the name of objectivity imply? In a way, it dehumanizes the behavior itself. A lie becomes defined by a certain pattern of oxygenated blood in the brain, completely decontextualized of the emotional dynamics that caused an individual to deceive in the first place.

Peeking into someone's brain is a powerful tool, and therefore highly coveted by security agencies in our current day and time. Some bioethicists are concerned that the fMRI lie detector may be abused. They believe some agencies might be tempted to use the test in situations where use of the test may infringe on individual privacy and autonomy. Furthermore, the false confidence in this technology could possibly blight a person's future, and many ask whether or not we are willing to take that risk. They wonder if the exaggerated reliability of the fMRI exemplifies how a technology can be spurred on by the convergence of basic science, the directing of research through funding, and certain groups who desire a particular technology. Essentially, this promoted reliability is pushing us further into equating the brain and the mind, physiology and emotion. What does this mean in terms of detecting deception among criminals? Stephen J. Morse, a professor of law at the University of Pennsylvania, believes that legally relevant neuroscience must begin with behavior, not with a picture of blood flows in the brain. "Brains don't kill people - people kill people."

What is the future of the fMRI lie detector?

To date, all fMRI lie detection studies have utilized small sample sizes with individuals who were asked to lie about very concrete facts such as the type of card they were holding. This is by no means a simulation of reality,one must account for different factors such as age, mental state, and, as Seelig pointed out, how good a liar one is. Pathological liars do exist, and these people may be so adept at suppressing the truth- our emotional baseline as postulated by Langleben- that they may never be caught. Whatever the degree of deception, the anxiety a suspect will face in a real-life detection situation is entirely absent from a lab setting. Furthermore, the fMRI does not account for substance abuse (prevalent amongst criminals), which can greatly alter BOLD levels. Given all these constraints, in order for the fMRI to have any credibility in a court, a consensus of scientists agreed that more testing is necessary, with larger subject groups and in ecologically valid circumstances in order to isolate these neurological markers of deception on a single-subject basis.

Where does this leave us in our search for the science behind lying? The need for transparency has intensified as we as a human race struggle to find security in the fragile and unpredictable economic, social and political constructions. An integral part of finding security is co-opting an identity that is morally and culturally acceptable. Some believe that an individual integrates both social and biological circumstances in constructing this identity. This co-opting of identities and categories is both bottom up (by a single individual) and top down (by a governing body). Detecting deception exemplifies this top down categorization, where designated figures of society- governments, lawmakers, security agencies, and corporations- are put in place to maintain the dynamics and identities of the consisting population. The fMRI lie detector may facilitate this process, but only to a certain extent. The scientists and the aforementioned social authorities must pose the question of how transparent the mind really is. Are we looking to create biological categories of "liars" and "truth tellers"? Moreover, is it ethically acceptable to create these types of stringent categories? The answers to these questions will determine the trajectory of the fMRI in lie detection, and more importantly, whether we will ever find a scientific explanation to an elusive human behavior that enmeshes us in a web of secrets, facts, forgotten memories and fantasy: Deception.


He published his findings in his treatise The Lie Detector Test on understanding physiological responses related to deception was published in 1938. Interestingly, he was also the creator of the character Wonderwoman', whose golden lasso- the Lasso of Truth- that forces people to tell the truth.

2 K.Alder. The Lie Detectors. Free Press, March 2007, p 34-36

3Dumit J, Picturing Personhood- Brain scans & Biomedical identity, Princeton University Press, 2004

4 Kittay.L . Admissibility of fMRI Lie Detection, Brookyln Law Review, p 12-13

5Langleben, D. Detection of Deception with fMRI- Are we there yet? , The British Psychological Society, Legal and Criminological Psychology (2008), 13, 1–9

6 Co-sponsored by Harvard and the McGovern Institute for Brain Research at the Massachusetts Institute of Technology (MIT))


• Websites






- Stanford Neuroethics & Law Blog





- Simulated Test experience

• Books/Journals/Scientific Articles:

- The Polygraph and Lie Detection, Division of Behavioral and Social Sciences and Education, National Research Council

- Commentary- The Future of Forensic Functional Brain Imaging,Daniel Langleben and Frank M. Dattilio, PhD

- How Honesty Testing Works, Miner. J and Capps. M, Quorum Books 2007