Forget lie detectors, brain waves could tell all

29 May 2017

Brain wave technology being tested in New Zealand is promising one of the biggest breakthroughs in forensic evidence since the use of DNA.

But some experts fear the technique hasn’t been tested properly, and are sceptical of claims of 99.9 percent accuracy.

Forensic Brain Wave Analysis (FBWA) - also known as ‘Brain Fingerprinting’ - is designed to read brainwaves to detect whether brains respond with recognition to certain information. It hinges on the premise that human brains react differently to unknown, and familiar, stimuli.

The US-based technology uses electroencephalography (EEG) to measure brainwave responses, known as P300.

The project here is being funded by the New Zealand Law Foundation and spearheaded by the University of Canterbury, in conjunction with the Brain Research Institute and Otago and Massey universities.

While testing in New Zealand is only at a preliminary stage, local researchers are looking to a future where it will be used as evidence in court.

Overseas, critics have highlighted various concerns, from the fact the method has only been tested by a private company and has not been replicated, to heavy scepticism around founder Dr Larry Farwell’s claims of 99.9 percent accuracy.

Those concerns are mirrored in New Zealand, with a University of Auckland cognitive neuroscientist warning of the pitfalls of taking research that hasn’t been replicated independently at face value.

However, local researchers working to bring the technology to New Zealand say they are confident.

Professor of Law at the University of Canterbury Robin Palmer is one of a core team of three testing the method.

He says if analysis is done properly, it’s very difficult to get a false result. Unlike polygraphs - also known as lie detector tests - the method isn’t designed to detect deception, but to identify whether people’s brains respond to information with familiarity or not.

“If you have information only a perpetrator would have of a crime scene … the known information would show as a peak [in monitoring the brainwaves] and the unknown information would flatline. You could say with some confidence that it’s in his brain. If he denies that knowledge, he’s got some explaining to do.”

Crucially, this means there would be no confusion between anxiety - often misidentified as a symptom of lying - and whether the person was telling the truth. Ideally, it puts people into one of two camps, having knowledge of the crime or not, Palmer says.

People are approaching the technology understandably tentatively in New Zealand, Palmer says, and no one wants to be the first to use it in court.

Other concerns include the influence of drugs and alcohol on recall, and the possibility of false memory implantations.

The team is 18 months into the research, and still has 10 scientific areas it wants to isolate for further study, including the impact of false memory.

While a local rollout is far-off, brainwave analysis has already been admitted as evidence in some criminal cases overseas.

The most well-known case in the United States may have been Harrington vs the State of Ohio in 2001, which admitted brain fingerprinting evidence, but ultimately did not decide the matter based on this evidence.

More controversially, an Indian court jailed a woman for life in 2008 after using FBWA evidence to convict her of murder. The judge held up as proof that the results showed the suspect's brain held “experiential knowledge” about the crime that only the perpetrator could have.

The decision was widely panned, with critics claiming there was no established marker for 'familiarity'.

Farwell, who runs the company Farwell Brain Fingerprinting, has openly rejected criticism, saying his techniques use well-established EEG technology to monitor brainwaves for recognition in specified contexts, and nothing else.

He touts the technology as a potentially powerful weapon against terrorism, saying analyses had been tested by the FBI, the CIA and the US Navy.

“There’s no such thing as ‘100 percent’ in science. I can’t say something’s 100 percent accurate, because nothing is.

“However, in every case so far, in hundreds of cases … we’ve got the correct answer,” he has told media in the past.  

His website describes brain fingerprinting as “a scientific method to discover the truth, identify criminals and terrorists, and clear the innocent with extremely high accuracy by measuring brainwaves”.

Neuroscientist and biomedical engineer Dr Richard Jones - another of the core group in New Zealand - says brain fingerprinting would provide a powerful local tool for forensic science.

However, he does acknowledge that - given it is almost impossible to replicate Farwell’s studies - accepting the accuracy claims demands an element of trust.

“We’ve had to take Larry [Farwell’s] results at face value … but there is no indication that there is anything wrong.”

Farwell was in New Zealand last year to speak about forensic brainwave analysis. As part of a presentation to the University of Canterbury, he tested brain fingerprinting on two law students.

“One of them had knowledge of a particular event in the last year and the other one didn’t, and the results were clear from this brain analysis who the so-called ‘guilty’ party was,” says Jones.

Jones says the analysis has had no false detections nor mis-detections to date, and if set up properly, it “never fails”.

However, University of Auckland cognitive neuroscientist Dr Reece Roberts says Farwell’s results aren’t mirrored in published literature, and 99.9 percent accuracy is unheard of in the field.

“This isn’t physics!”

While he concedes the science isn’t “entirely bogus”, Roberts says Farwell’s research is for a private company, so outcomes should be regarded with scepticism.

“We should be wary about saying that you can tell innocent people from guilty people to a very high degree.

“The effects certainly aren’t as strong as [Farwell] is claiming. There is some evidence but it’s nowhere near 100 percent,” Roberts says.

“My main concern is this is something that could be used in principle in a court of law but it would need to be replicated quite extensively by independent research groups. Until that happens, the responsible thing to be is to be sceptical.” 

Professor Peter Rosenfeld, of NorthWestern University in Chicago, is a longtime critic of Farwell’s method.

A major concern for him is that “recognition” can be foiled by associations of other kinds. In one of Farwell’s tests, the brain’s “recognition” of the hammer as a murder weapon could be interpreted as just that. But a subject who had a different type of connection or association with a similar hammer might see it as a piece of nostalgia - a hammer not unlike one a beloved grandparent might have kept in a shed of the subject’s youth, thus sparking a different kind of recognition - but recognition all the same.

He points to 2012 peer-reviewed research co-authored by Emanuel Donchin - considered by many in the field as the world's leading expert on the P300 brain wave - which found the P300 is activated not just by familiarity, but “any event that violates the subject’s expectancies”. Rosenfeld backs the findings, which say there is no simple one-to-one relationship between the P300 and memory.

In other words, it would seem test results can be tampered with - incidentally or intentionally.

The same research paper claims of the three laboratory studies on brain fingerprinting Farwell claims results of 99.9 percent accuracy for, only one was published in a peer-reviewed journal. The one published claimed to have tested 40 participants, while in fact it examined 20 participants twice, leading the study to label Farwell’s findings as “misleading”, and a misrepresentation of the scientific status of brain fingerprinting technology.

Rosenfeld, who has been working in the same field for 40 years, replicated some of Farwell’s studies himself, and achieved results closer to the 80-82 percent mark.

Much like the evolution of DNA evidence in courts, it would be a long time before the technology could become mainstream in New Zealand, even after it had been fully approved.

The equipment’s prohibitive cost is one of the things the team of researchers must find a solution to if they want to see it rolled out here, and while some funding has been secured for initial testing, they have a long way to go.

“We’re now looking at putting in a bigger, second application [to the Law Foundation] but there’s no guarantee that we’ll get that,” says Jones.

“Things are a bit on hold until we secure the funding.”

Dr Reece Roberts is a research fellow in the Department of Psychology, in the University of Auckland's Faculty of Science.

Used with permission from Newsroom, Forget lie detectors, brain waves could tell all published on Monday 29 May 2017.