Features
Prof Sounds Alarm on Forensic Tool
Neuroimaging less accurate than believed, study finds
BY DEIRDRE HEALEY
Technology that can monitor brain activity is moving into the world of forensic investigation faster than it should, says Prof. Dan Meegan, Psychology.
In a first-ever analysis of the research on “Brain Fingerprinting” and other neuroimaging techniques, he has found that the technology hasn't been proven reliable enough to determine whether someone is guilty or innocent of a crime.
“There's a huge amount of interest in this technology coming from the legal community, and if it's being used to determine whether people are innocent or guilty, then it's important that we find out if these techniques are reliable,” says Meegan, who specializes in memory-related research.
“From a scientific perspective, I can tell you that these techniques aren't ready.”
Brain Fingerprinting is currently being used in criminal investigations in the United States and has even been accepted as evidence in court.
The technology involves attaching electrodes to a suspect's scalp and then monitoring how the brain reacts to certain stimuli, such as photos or words, related to a crime scene that only the person who committed the crime would know. The brain produces different electrical signals based on whether the person recognizes the stimuli or not. These signals are captured by the electrodes and analyzed.
Similar testing can be done with magnetic resonance imaging technology, where blood flow in the brain is used to determine whether the brain recognizes — or has memory of — certain stimuli.
Investigators and attorneys then use the test results as evidence to support a person's guilt or innocence.
This technology received significant media attention after the attacks on the World Trade Center because it was viewed as a potential tool for identifying people who were planning terrorist attacks or had knowledge of future attacks, says Meegan. At the time, he assumed critics of the technology would eventually come forward, but that never happened.
“After watching all the media publicity this technology received, I immediately thought of some of its limitations. I kept waiting for scientists in the field to step forward and argue that more work needs to be done before this technology can be used in the field, but that never happened, so I decided to take it upon myself.”
Meegan's research, which was published in the January issue of American Journal of Bioethics, found that the level of accuracy of neuroimaging techniques is not as high as believed.
“There is evidence of a substantial risk of the technology failing to catch people who are guilty, so results that find a person innocent are completely meaningless,” he says.
He also found that a majority of the research conducted on the accuracy of this technology was done in a lab setting, where the researcher has more control over the circumstances and subjects than would be possible in a real criminal investigation.
“With an actual investigation, there are a number of factors that can influence a person's ability to recognize — or have memory of — certain details.”
In some cases, weeks or even months will pass before a suspect is identified, he says.
“Researchers don't wait months to do tests, and this time lapse is important to consider because it can hinder people's memory of the crime and their ability to recognize certain stimuli.”
In lab settings, researchers can control which stimuli they want a subject to remember by showing it to the person in advance of the test. But when a person is committing a crime, he or she isn't often taking mental notes of the surroundings and may not be able to remember certain details of the crime, says Meegan.
In addition, suspects can be under the influence of drugs when committing a crime, which can impair their ability to remember details.
There's also an element of high emotion when a crime suspect is taking the test compared with a participant in a lab study, and this emotion can affect results, he adds.
“There needs to be more research done involving circumstances that mimic a forensic investigation before neuroimaging can be established as a reliable tool. Until then, scientists need to put the brakes on using this technology in actual investigations.”