Experts say a functional MRI is more accurate than a polygraph. However, more research needs to be done before judges will allow the tests to be used as evidence.
Searching for the truth is the very definition of what happens in criminal court.
Forensic science has long assisted in that quest.
However, other scientific tools — polygraphs, brain scans, and functional magnetic resonance imaging (fMRI) — remain largely inadmissible as evidence of guilt or innocence.
Some medical experts believe that could change if larger trials are conducted outside the laboratory in real-world conditions, using strict protocols that yield reproducible results.
Dr. Daniel D. Langleben is one of the leading researchers in the field of lie detection. He is associate professor of psychiatry in the Perelman School of Medicine at the University of Pennsylvania and a staff physician for the Philadelphia Veterans Administration Medical Center.
Does Langleben see a future use for fMRI in court cases?
“Yes,” he told Healthline. “But what stands between this being an educated answer with data behind it, and an educated guess, is we need to have larger-scale trials that test real-world situations under controlled conditions. Until this happens, my answer would be a guess.”
“Right now,” he continued, “we have the polygraph, which does have a level of accuracy significantly above chance. There are some people who would say that the polygraph is 100 percent accurate. But the literature as a whole, including the report by the National Academy of Sciences, points to a number in the 75 percent range. So, the polygraph is already pretty good but not good enough for real life, meaning clinical applications. If the fMRI could improve on that, it means there is a way forward.”
An fMRI was at the center of a high-profile murder case in Maryland.
The defendant, Gary Smith, a former Army Ranger with five combat tours in Iraq and Afghanistan, was on trial on charges of killing his roommate
Smith’s attorney hoped that his client’s fMRI would prove he was telling the truth. The judge presiding in the case said he found the fMRI “fascinating” but refused to admit it as evidence.
Langleben and Jonathan G. Hakun, PhD, an assistant teaching professor in psychology at Penn State, published
“We showed a 12 to 17 percent difference between polygraph and fMRI, in favor of fMRI,” Langleben said. “[An] fMRI could be used for lie detection and it could be better than polygraph. But it would not answer one key question: Will it ever be good enough for legal implications? Because there we require a completely different level of accuracy.”
The polygraph, introduced more than 50 years ago, monitors a person’s electrical skin conductivity, heart rate, and respiration during a series of questions.
The assumption is that upward or downward spikes in those measurements indicate that the person is lying.
While polygraph results have been judged to be inadmissible as legal evidence in most U.S. jurisdictions, they have been used for nearly 30 years in the business world as a device for pre-employment screenings. Polygraphs are also used widely in government background checks and security clearances.
“Polygraph measures reflect complex activity of the peripheral nervous system that is reduced to only a few parameters, while fMRI is looking at thousands of brain clusters with higher resolution in both space and time,” Langleben said. “While neither type of activity is unique to lying, we expected brain activity to be a more specific marker, and this is what I believe we found.”
However, some legal experts remain skeptical about brain scans as a lie-detection tool.
Henry T. Greely, JD, a professor of law at Stanford University in California and director of the Stanford Center for Law and the Biosciences, said any single study “needs to be viewed skeptically, no matter how good the investigator.”
“If five different teams replicated the Langleben study, I would feel much better about it, in part because it would have involved more than only 28 people,” he told Healthline. “Even then, lies told by people who know they are research subjects, and are following instructions to lie, may look very different from lies in real life.”
“That’s a very hard problem to solve,” Greely added. “We can’t go around arresting people to make them take an fMRI test to test ‘real’ lying. In any event, ‘significantly better’ than the polygraph is not very good. In almost every U.S. court, it is not good enough to be admitted, and most experts think it should not be used as often as it is outside of court. That’s the most important bottom line: better than the polygraph, even if true, isn’t good enough to be used for important decisions.”
Greely said judges in all the cases where the evidence has been presented have rejected fMRI after hearing expert witnesses because its results aren’t proven to be sufficiently accurate and the tests did not follow any well-established protocols.
In addition, he said, “the evidence would eat up too much time and cause too much possible jury misunderstanding to be worthwhile in light of its very dubious value.”
A radiologist agrees with Langleben on the need for advanced testing of fMRI, outside of the laboratory.
Dr. Pratik Mukherjee is a professor of radiology and bioengineering at the University of California, San Francisco (UCSF) and director of the Center for Imaging of Neurodegenerative Diseases at the San Francisco Veterans Affairs Medical Center.
“Testing could possibly be conducted in real-life legal cases but would have to be done under strict scientifically rigorous conditions,” he told Healthline. “Since this would constitute research, the admissibility of any results in court would be questionable until the testing is fully validated. This is similar to the ethical barriers to using the results of research studies for clinical practice in medicine.”
Mukherjee said certain standards must be met before the doubts and objections to the admissibility of fMRI in court cases can be overcome:
- Accuracy. It must have acceptably low rates of false positives and false negatives.
- Reliability. It must have acceptably low rates of failure.
- Generalization. Does it work in people of all ages and IQ levels, those with mental illness, those under the influence of psychoactive substances, and those with prior head injury, stroke, dementia, and so on?
- Robustness to countermeasures. Simply moving your head slightly during the scan is enough to degrade any fMRI.
“Better brain imaging methods are needed, and a lot more scientifically rigorous testing, including under real-world conditions,” Mukherjee said. “Even much of the current scientific literature using fMRI for academic neuroscience research is suffering from a failure of reproducibility. The emphasis now is on improving fMRI methodology to produce more reliable and reproducible results.”
How will Langleben test fMRI outside of the laboratory?
“Very similar to the way they use polygraph in Japan,” he said. “Someone with appropriate expertise will study the case and assemble a ‘forced choice’ questionnaire with questions that have clear yes/no answers that maximize the difference between a likely perpetrator and the person being tested.”
Results of the analyzed data will have a quantitative estimate of the “effect size” — the strength of the difference between lie and truth, he said.
Langleben said he knows why courts continue to resist the admissibility of fMRIs:
“Legitimate concerns about lack of data on the ‘error rates’ of this approach under ‘real life’ circumstances,” he said, “unfounded fears of being supplanted or even replaced by the new technology, and irrational fears of having one’s mind looked into. Basically, the good old Freudian ‘resistance.’”
Judges use the established Frye (1923) and Daubert (1993) standards to determine whether they will allow polygraph or fMRI results to be admissible in their courtrooms.
A court that applies the Frye standard must determine whether the method by which the evidence was obtained was generally accepted by experts in a specific field.
With Daubert, a trial judge makes a preliminary assessment of whether an expert’s scientific testimony is based on reasoning or methodology that is scientifically valid and can be applied properly to the facts of the case.
The Frye standard has been abandoned by many states and the federal courts in favor of the Daubert standard, according to the website of the Legal Information Institute, based at the Cornell Law School.
Meanwhile, Joel Huizenga, chief executive officer of Truthful Brain Corp. in California — who conducted the fMRI on ex-Army Ranger Gary Smith — is working on another murder trial through the Innocence Project.
Huizenga sees fMRI as a valuable tool to measure whether a defendant tells the truth.
“The National Academy of Sciences came out with a report that concluded none of the technologies presently being used as forensics by the court system had been shown by scientific methodologies to work or be accurate, except for DNA testing,” Huizenga told Healthline.
“All these others (fingerprints, etc.) were grandfathered in without any proof that they worked, and presently could not pass the Frye or Daubert tests for admission to the court system for use,” he added.
The United States sends people to death row with eyewitness reports, which have been shown to be 65 percent accurate when they are done the in the traditional way, Huizenga noted.
“If you give the images one by one and tell the person that the perpetrator might not be in the list, which is a new methodology, the accuracy goes up to 75 percent,” he said. “So, thinking the court system is about accuracy is ridiculous. It is about power, and it is definitely anti-science in general, since science takes power away from the workers in the legal field to do more of what they want to do.”
“Currently, there is a power struggle between science and law. Law is winning big, at the expense of our population,” he added.
Greely notes that DNA evidence for identification is a “much more scientifically easy process.”
“But it took two reports by the National Academy of Sciences and an FBI program to create protocols for its use,” he said, “and to accredit crime labs to do that testing before it was widely accepted. If fMRI-based lie detection is ever very useable — and I put the odds at about 50/50 in the next 10 to 20 years — similar things will have to happen.”
Andrew Jezic, the Maryland criminal defense attorney for Gary Smith, introduced his client’s fMRI at his second trial in 2012. The judge did not admit it.
Smith was found guilty twice and his conviction was overturned twice, Jezic said. Smith recently made an Alford plea.
“It was not an admission of guilt,” Smith told Healthline. “I pled guilty to involuntary manslaughter and reckless endangerment, but I maintain my innocence. I had lost nearly a decade of my life — six years in prison and three years of house arrest. The Alford plea gave me time.”
The next step in the Alford plea process will be a reconsideration hearing in front of a judge, but Smith will have to wait 18 to 24 months before getting that hearing.
Smith is now finishing college, works as Jezic’s law clerk, and plans to attend law school.
Jezic calls fMRI a “fabulous tool.”
“The fact that someone is willing to submit to it is a factor in and of itself,” he told Healthline. “It takes courage to submit to an fMRI when you are told ahead of time that this is not something you can fake, and is not something you can read about on the internet to help you pass this test. If somebody’s willing to do this and goes through with it, that’s a pretty significant indicator that the person believes they are innocent.”
Jezic said fMRI is a long way from being admissible, but is “interesting” for all parties involved.
“If a guy takes an fMRI and fails miserably, that will probably affect the mentality of the defense attorney and the prosecutor,” he said. “If the person passes the test, it may not affect the defense and prosecution, because they’re not going to believe anything about it. But it’s a demonstration of courage and of the person’s absolute conviction that he or she is innocent.”