"Brain Fingerprinting" as Evidence of Criminal Guilt?

Via Elizabeth Joh, I learned about the company, Brainwave Science which specializes in "brain fingerprinting" technology. According to the company, brain fingerprinting can assist in determining whether a suspect is guilty of a crime. From the company's website:

Once a suspect is in custody, Brain Fingerprinting can accurately establish whether or not the suspect has knowledge of the crime. This saves time and money on investigations, while keeping dangerous criminals off of the streets. 

Conventional fingerprinting and DNA testing match physical evidence from a crime scene with evidence found on the perpetrator. While these are tested and accepted methods, they are only available in roughly 1% of crimes. It is estimated that Brain Fingerprinting would be applicable in 60%-70% of these same cases, thus having the potential to make a profound affect on the criminal justice system. 

The advantages to Brain Fingerprinting are that it offers higher accuracy than other techniques, is non-invasive and has been ruled admissible in the US court.

There are some . . . interesting claims being thrown around in these paragraphs (I want some more detail on that fingerprinting and DNA evidence only being available for 1% of crimes). But I was particularly intrigued by the last paragraph, which notes that brain fingerprinting has been ruled admissible "in the US court" (whatever jurisdiction that's supposed to mean). The website also claims that the technology has been used to obtain convictions.

I was intrigued by these claims, since the Brainwave Science is making some pretty outrageous far-reaching claims about the power of its technology. And I have reason to doubt the company's claims of legal success because I think that this sort of evidence would raise serious evidentiary concerns.

In the rest of this post, I address Brainwave Science's claims. I determine that some of the company's claims are highly misleading, while others are a bit more borderline. In any event, serious questions remain as to whether this sort of evidence would be admissible in trial.


Brainwave Science's Claim that Brain Fingerprinting has Been Admitted in Court

Brainwave Science claims that brain fingerprinting evidence has been admitted in court. Following the links on the website clarifies that the company is referring to an Iowa case involving the defendant, Terry Harrington.

More information on the case and the brain fingerprinting technology at issue can be found here. Brainwave Science uses Harrington's case to paint a rosy picture of its technology and its admissibility:

In 1977, at the age of 17, Terry Harrington was arrested for murder. He was convicted and spent over half of his life in prison. Twenty-two years after his conviction, Dr. Lawrence Farwell used Brain Fingerprinting testing to show with a 99.9% statistical confidence level that the record stored in Harrington's brain does not match the crime scene and does match his alibi. The testing showed that significant details of the crime are not stored in his brain. On February 26, 2003 the Iowa Supreme Court reversed his murder conviction and ordered a new trial. In October 2003, the State of Iowa elected not to re-try Mr. Harrington and released him from prison. 

In the Brain Fingerprinting tests, Harrington's brain did not emit a MERMER in response to critical details of the murder, details he would have known if he had committed the crime, indicating that this information was not stored in his brain. In a second Brain Fingerprinting test, one that included details about Harrington’s alibi, Harrington's brain did respond with a MERMER, indicating that his brain recognized these events. The details used in the second test were facts about the alibi that Dr. Farwell obtained from official court records and alibi witnesses. 

"It is clear that Harrington's brain does not contain critical details about the crime," said Dr. Farwell. "His brain does, however, contain critical details about the events that actually took place that night. We can conclude scientifically that the record of the night of the crime stored in Harrington’s brain does not match the crime scene, and does match the alibi.” 

The science of Brain Fingerprinting worked to exonerate Terry and can now be used to increase public safety across all areas of crime.

This looks pretty good for the brain fingerprinting technology. Based on this statement of the case, it looks like the Iowa Supreme Court examined this new evidence and determined that it warranted a reversal of the opinion.

But that's not what happened.

It is true that the Iowa Supreme Court mentioned the new brain fingerprinting evidence in its opinion in Harrington v. State. But that evidence did not at all factor into the court's determination that Harrington's conviction should be overturned.

While Harrington mentioned the evidence in his request for a new trial, the Iowa Supreme Court determined that the prosecution had suppressed multiple police reports at trial. These reports contained exculpatory evidence that indicated that another party may have committed the crime. Because the prosecution had withheld material, exculpatory evidence, the Iowa Supreme Court concluded that Harrington's due process rights had been violated under Brady v. Maryland.

At most, the evidence may have been admitted in the district court case below (since the Iowa Supreme Court mentions "testimony" concerning brain fingerprinting technology in footnote six of its opinion). But that is not entirely clear from the limited discussion of the technology in this opinion, and there is no other mention of brain fingerprinting technology in any other opinions following the proceedings in which Harrington was involved.

Just because the Iowa Supreme Court mentioned the brain fingerprinting technology in its opinion reversing Harrington's conviction does not mean that the technology factored into the Court's conclusion. Here, the evidence played no part in the Iowa Supreme Court's reasoning or conclusion.

By claiming that brain fingerprinting technology "worked to exonerate Terry" and by being very selective with the facts of the Iowa Supreme Court case, Brainwave Science's use and characterization of the Harrington case as a success story is highly misleading.

Brainwave Science's Claim that Brain Fingerprinting Helped Obtain a Conviction

Brainwave Science also claims that brain fingerprinting technology brought a "serial killer to justice." This claim should also be taken with a grain of salt, although I admit that it is not as blatantly misleading as the company's earlier claims.

The company describes the case of James Grinder, a man suspected for 15 years of murdering a victim, Julie Henton. Over the lengthy investigation, Grinder apparently gave various conflicting accounts to officers about his involvement in the crime. Eventually, the Sheriff sought to use brain fingerprinting technology on Grinder. From the website:

Computer analysis of the Brain Fingerprinting test found, with a statistical confidence level of 99.9%, that the specific details of the crime were recorded in Grinder’s brain as “information present”, which means that record stored in Grinder’s brain matched the details of Julie Helton's murder. 

Result: Brain Fingerprinting Testing Traps Serial Killer in Missouri, USA 

Following the test results, Grinder faced an almost certain conviction and probable death sentence. Grinder pled guilty to the rape and murder of Julie Helton in exchange for a life sentence without parole. He is currently serving that sentence. In addition, Grinder confessed to the murders of three other young women.

There is no information on whether Grinder objected to the use of the brain fingerprinting technology. Moreover, it is unclear how strong the case against Grinder was. While the case may have been weak given the years that had passed since the murder, Grinder had also apparently been giving conflicting accounts of involvement, which would have put in him on the radar as a serious suspect.

Whatever the circumstances, Grinder apparently felt that the case against him was strong enough to warrant a guilty plea. Because he took a plea, however, whether the brain fingerprinting would have been admissible in trial remains an unanswered question.

If the brain fingerprinting were truly a crucial part of the case against Grinder, Brainwave Science's claim that Grinder faced "an almost certain conviction" hinges on the brain fingerprinting evidence's admissibility in the criminal trial. Brain fingerprinting raises some serious evidentiary questions, and I conclude this post by exploring these questions.

Is Brain Fingerprinting Evidence Admissible?

Courts generally apply one of two tests to determine whether expert testimony is admissible. A minority of courts apply the "Frye" test -- named for the 1923 District of Columbia Circuit case of Frye v. United States -- which requires expert evidence to be sufficiently well-established to be generally accepted in its field. This is generally recognized as taking a fairly narrow view of the admissibility of expert testimony.

The federal courts and many states evaluate the admissibility of expert testimony under the rule established in Daubert v. Merrell Dow Pharmaceuticals, Inc. The Daubert test allows a broader range of expert evidence, and requires the court to look to several factors when evaluating whether evidence is admissible. These factors include whether the expert's methodology has been or can be tested by the scientific method, whether it has been subjected to peer review or publication, the method's potential rate of error, and whether the method has been generally accepted by other experts in the field.

In this paper in The Scientific Review of Mental Health Practice, J. Peter Rosenfeld raises some serious concerns with brain fingerprinting's methodology and reliability. Rosenfeld's criticism raises serious doubts over brain fingerprinting's admissibility under even the comparatively liberal Daubert standard, noting that the technology's main proponent, Lawrence Farwell, has serious problems defining his methodology and how it can be tested, and that there is a dearth of published studies on brain fingerprinting methodology.

Rosenfeld's paper is dated, however, having been published in 2005, and Farwell has since fired back with this paper specifically replying to Rosenfeld's criticism and several other papers here and here detailing brain fingerprinting's accuracy.

While there is quite a bit to sort through between all of those papers, I am on vacation in Iowa and I left my neuroscience-expertise hat back in California, so I unfortunately will not be able to parse through the science in any detail. But it looks like there is enough controversy surrounding brain fingerprinting methodology that there would be some problems admitting that type of evidence under both the Frye and Daubert tests.

Two final observations. First of all, recall Brainwave Science's claim that James Grinder faced " an almost certain conviction" after brain fingerprinting indicated his guilt. While Farwell has published some more papers on his methodology in recent years, Grinder underwent brain fingerprinting in 1999, where there was a serious lack of available information and published scholarship on brain fingerprinting and its accuracy. There may be a somewhat colorable argument to admit brain fingerprinting evidence today. But back in 1999, it would have been far more accurate to conclude that there was an almost certain chance that the brain fingerprinting evidence would not have been admissible.

I would like to conclude by emphasizing that Brainwave Science is claiming that brain fingerprinting has "an accuracy rate of over 99%." Say what you will about the state of the science -- I find it extremely hard to believe that a device that reads brain waves can determine whether somebody is lying or telling the truth with what is essentially 100 percent accuracy.