CIA double agent Aldrich Ames managed to pass polygraph tests, designed to find traitors, even as he was selling out other US agents to the Soviet Union. Polygraphs — conventional lie detector tests — are based on emotional responses to stress such as increased heartbeat and blood pressure, and excessive sweating. Responses that can be faked.
But whatever the outward response, the brain will always answer honestly. Discovering this answer from involuntary brain activity is where brain fingerprinting (BF) comes in. BF taps into the specific information stored in a person’s memory, providing a scientific solution to the problem of identifying criminals and trained terrorists. This scientific determination of guilt or innocence has been ruled admissible in US court, meaning this new branch of forensic science has the potential to revolutionize the whole justice system. Indeed, the revolution has already started.
In February 2004, a Brain Fingerprinting test showed that Jimmy Ray Slaughter's brain did not contain a record of salient features of the crime of which he had been convicted.. He is on death row in Oklahoma for the murder of his girlfriend Melody Wuertz and their daughter Jessica. Slaughter’s fate is not yet known. But, the implications of the technology will reach much further than the criminal world. BF creators Brain Fingerprinting Laboratories, based in Seattle, say the same technology will be used to speed to market drugs for brain disorders including Alzheimer’s disease (AD) and Hutchinson’s, help identify fraud, and strengthen security in areas such as visa applications and protection of classified information.
Farwell measures brain-wave responses of a person looking at words or pictures displayed on a computer screen using a headband with built-in electrodes The Gazette/Buzz Orr
The most detailed knowledge of any crime is locked away in the brain of the person who committed it. BF provides a method by which these invisible clues can be tapped to determine, with scientific certainty, whether the story a suspect is telling matches what is stored in his or her brain.
BF test results suggest that Slaughter’s story and the information in his brain match. His claim to have no knowledge of the murder scene was proven to be true, enough to question the validity of the guilty verdict and have him granted a stay of execution. The same technology could soon be available to establish the guilt or innocence of suspects at a much earlier stage.
‘I envision that over the next 10–20 years police officers and investigators throughout the world will be trained as part of their regular law-enforcement education to record the elements of a crime scene for use in BF tests,’ says Drew Richardson, a former US Federal Bureau of Investigation (FBI) agent and scientist, who joined Brain Fingerprinting Laboratories. ‘Up to 70% of major crimes would be appropriate someday for BF technology,’ he says. Conventional fingerprinting and DNA is only available in 1% of cases. Brain Fingerprinting Laboratories hold three patents for BF testing in the forensic area.
The technique is not for interrogation. It does not require any questions or any answers. It reveals objectively whether information is present in the brain, regardless of whether any false or truthful statements are made by the subject. The brain does the talking.
Hidden information is uncovered by measuring brain-wave activity in response to crime-related pictures or words. Changes in brain waves allow a determination of ‘information present’ or ‘information absent’ regarding specific details of a crime in a specific brain. ‘We cannot download the contents of the brain,’ says BF inventor Laurence Farwell of Brain Fingerprinting Laboratories. ‘But what we can do is determine whether the suspect recognizes details significant in the context of a crime.’
Presented with details of the crime, the guilty person cannot help but elicit an involuntary, but detectable spark of recognition in the brain. This response is automatic, so there is no way to suppress or fool the system. ‘If an elephant were to walk into the room, you may not respond overtly, but your brain cannot help but recognize that it is an elephant. There is always the “ah-ha” response in the brain,’ says Farwell.
Find the MERMER and you have found the murderer
The ‘ah-ha’ response is characterized by specific, measurable brain response known as a MERMER. A MERMER will only be emitted by the brain of the perpetrator, with details of the crime in his brain, and not by an innocent suspect who does not have this record in his brain. Find the MERMER and you have found the murderer.
A MERMER is a part of the brainwave observed in response to familiar information. When the brain recognizes something, the memory centres are stimulated. The neurons fire synchronously, eliciting characteristic changes in brain activity. It is these changes, which can be measured using electrodes, that investigators look for when trying to determine whether someone recognizes a particular piece of information.
When subjected to a rare, but meaningful stimulus, increased neuron activity results in an increase in voltage, typically within 300–1000msec after the stimulus, called a P300. This is called an event-related potential (see Box, above). For example, if a subject is exposed to a series of random names and occasionally one of those names is the subject’s name, a P300 response is evoked.1,2 The utility of the P300 in detection of deception was recognized as early as 1988, when it was shown that it could be used to identify college students concealing guilty knowledge of having stolen something.3 However, the P300 has only a 87.5% success rate in revealing the presence of relevant information.
Farwell’s test is based on the discovery that the P300 is only a subcomponent of a more complicated response called a MERMER, which is elicited when a person recognizes and processes a stimulus that is particularly noteworthy to him/her.4 The MERMER, memory and encoding related multifaceted electroencephalographic response, includes the P300 and another longer latency, electrically negative subcomponent with a latency of up to two seconds post-stimulus. In other words, a positive wave followed by a negative one. Tests using the MERMER produced no false negatives or positives and no indeterminates.4
When details of a crime are known to the suspect, a MERMER will be detected. A MERMER will not occur in an innocent subject.
However, there is one major drawback. Although it is possible to determine whether information is present in the brain, it is not possible to say why it is there. This is why investigators have to find information that only the guilty person will have. This means eliminating any details that may have come out during the trial, for example, to which an innocent person may have been inadvertently exposed.
There are about a billion neurons in the brain that communicate with each other using electrochemical signals. The ongoing changes in these signals, which are measured using scalp electrodes, are recorded as continuous changes in voltage over time, called the electroencephalogram (EEG). Buried in the EEG are signals that reveal information about brain processes. These signals are detected by timing changes in the EEG with the onset of events such as listening to a sound or viewing a picture. The resulting activity is called an event related potential (ERP), which is distinguishable from background brain activity. The ERP can be broken down into several basic components represented as positive or negative fluctuations in the ERP waveform. Components that occur prior to 100ms are thought to reflect information processing in the early sensory pathway, for example the auditory neural ERP stems from neuronal impulse traveling from the cochlea through the auditory brain centres. Longer latency ERP components include P1, P2, N1, N2, N400 and P300 components. These are named by the polarity (P for positive) and either their ordinal position (P1 is the first positive wave) or their latency after onset of stimulus (N400 is a negative fluctuation peaking at 400msec from onset of stimulus). Generally the components occurring before 250msec are thought to reflect late sensory and early pre-conceptual processes, whereas those after 250msec are thought to reflect higher level cognitive processes such as memory or language.
‘Brain fingerprinting could have significant commercial potential within market segments of interest to Lilly,’ Christian Fibiger, Lilly Research Laboratories
Initial lab testing of BF technology was funded by the US Central Intelligence Agency (CIA) and the FBI. One of the first studies involved testing which people in a group were FBI agents, by looking for a MERMER in response to words and phrases that only FBI agents would recognize. A headband with sensors is placed on the subject’s head and a series of words and pictures, some of which are relevant and some are not, are flashed on a computer screen (see Box, below). The brain waves produced in response are recorded and analyzed to determine whether there is an involuntary spark of recognition to any of this information. The FBI agents were identified with 100% accuracy. Similar tests with US navy doctors were equally successful. BF was also used to identify who, of four volunteer test subjects, was responsible for the theft of anhydrous ammonia (used to make methamphetamine) from a farm in Fairfield, Iowa. Each subject was exposed to information relevant to the crime to determine which one of them exhibited the ‘ah-ha’ response. The guilty party, who had already confessed and done jail time, was identified.
Rigorously conducted lab tests are one thing, but looking into the mind of a real killer is a different story. Farwell had the opportunity to do this when sheriff Robert Dawson called on Brain Fingerprinting Laboratories to test JB Grinder who, for 15 years, had been the chief suspect in the vicious murder of a young girl in Missouri. To identify the murderer, the investigators had to uncover information that only the real murderer would have, and establish that Grinder could not have that in- formation for the wrong reasons
(for example, because somebody told him or he had seen the crime scene innocently). Tests showed that the record stored in Grinder’s brain matched the crime scene. Faced with certain conviction and a probable death sentence, Grinder pled guilty in exchange for life without parole and confessed to three previously unsolved murders of young women. BF was also used to exonerate a man who said he was wrongly convicted of murder. Terry Harrington (a black man) was found guilty of murdering a (white) security guard in Iowa in 1978. BF tests on Harrington revealed that the information in his brain did not match the crime scene (see Figure 1). The brain response to the probes (blue line) is the same as the response to the irrelevants (green line). Harrington’s brain response to his alibi show the information stored in his brain matches his alibi (see Figure 2). The judge ruled that the BF test results ‘meet the legal standards for admissibility in court for scientific evidence,’ and the murder conviction was reversed in 2003. In the Slaughter case, BF testing proved that the suspect had no knowledge of the baby’s bullet wounds, where the woman was killed or details of the stab wounds inflicted. Preliminary analysis suggests that there is a 99% chance that his brain record does not match what he is convicted to death row for having committed, says Farwell. All Slaughter’s appeals had run out, but Slaughter's attorneys claim that the BF test results are sufficiently scientifically rigorous to warrant a stay of execution.
Multiple-choice test for the brain
Three kinds of information are used to determine whether a subject has specific crime-related information in his brain:
• Targets: information the subject definitely knows; this can be ensured by telling the subject before the test starts.
• Irrelevants: information that subject definitely does not know; this can be ensured by simply making up the information.
• Probes: information relevant to the crime or situation, which the subject may or may not know.
The response of the brain to information is measured using a headband with electrodes. Target information elicits a ‘yes’ response or a MERMER. This is used as a control. Irrelevant information will not elicit a MERMER. A MERMER in response to probe stimulus indicates recognition or the presence of certain information.
MERMER testing will also be used by pharmaceuticals companies to speed to market new drugs for brain disorders such as AD. Brain Fingerprinting Laboratories has already had discussions about a possible collaborative venture with drug giant Eli Lilly.
‘Clearly there exists in AD and other cognitive disorders a need for both improved drug-efficacy testing protocols and accurate treatment monitoring,’ says Christian Fibiger, vice president of neuroscience discovery research and clinical application at Lilly Research Laboratories in Indiana. ‘[BF] could have significant commercial potential within market segments of interest to Lilly,’ he says.
Brain Fingerprinting Laboratories has already patented a technology that for the first time can measure how memory and cognitive functioning of AD patients are affected by medications on a short-term basis. And even the rudimentary first-generation test protocol has been shown to detect AD as accurately as magnetic resonance imaging, positron emission tomography and routinely used subjective testing. The new test only takes 30 minutes and is based on a personal computer. Similar test protocols usable with the same equipment for other neurological disorders, such as Hutchinson’s disease, are also being developed.
This technology will speed up development and US Food and Drug Administration approval of new drugs by improving testing at clinical trials, aiding in the early diagnosis of AD, and providing physicians with a tool to determine the short and long-term effects of a particular drug.
This technology is expected to have particular application in the development of a class of drugs called gamma secretase inhibitors. These drugs reduce the accumulation and deposition of beta-amyloid plagues, insoluble protein structures that cause memory problems when they accumulate in the brain. MERMER testing can be used to evaluate the efficacy of the drug in slowing the progression of memory loss caused by amyloid plagues. Currently, there are 19 drugs under development to slow the progression of AD or prevent it.
In AD testing, technicians present words, phrases and images that are both known and unknown to the patient. The brain responds, and the technician can use this response to measure deterioration of memory and also deterioration of the speed, accuracy and efficiency of the cognitive processes. The number of tests for individual patients would range between four and 12/year. The results would also be used by physicians in a clinical setting to evaluate the medications already being prescribed, to decide whether they should be changed. A commercial testing product is expected to be available within 18–24 months.
With the average age of the population increasing, the number of people with AD is expected to increase nearly 300% by 2050. The US market alone has the potential for AD systems and services in excess of $2bn/year. The market for BF equipment is estimated to be 60,000 units at a cost of between $15,000 and $25,000.
Another potential application of the technology is identifying trained terrorists before they strike, including those in long-term ‘sleeper’ cells. BF technology can be used to detect records in the brain of crimes in the planning or association with terrorist groups. Tests will use information that only terrorists or gang members would have access to. This might include details of training camps or locations specific to a particular group or gang. The results of the tests will likely be used to make decisions over who can be safely issued a visa or given leave to enter a foreign country.
1 Donchin et al, Behavioural Brain Sciences 1988, 11, 357
2 R Jr Johnson, Advances in psychophysiology 1988, 2, 69
3 Rosenfeld et al, International Journal of Neuroscience 1988, 24, 157
4 Farwell et al, Journal of Forensic Sciences 2001, 46, 1
Copyright 2012 Society of Chemical Industry
- all rights reserved.