## Facts

- Lorne Allan Semrau, a licensed psychologist, was indicted for a scheme to defraud government health-care programs, including Medicare and Medicaid, by submitting false claims for payment.
- Semrau intended to defend in part by arguing he did not act with intent to defraud when claims were submitted.
- Semrau’s counsel contacted Dr. Steven Laken, the president and CEO of Cephos Corporation, a company developing functional magnetic resonance imaging (fMRI) for use in lie-detection or “truth verification.”
- Semrau underwent two fMRI examinations arranged through Cephos, and the results were interpreted by Laken as showing Semrau was truthful when denying intent to defraud.
- Laken proposed to testify that the Cephos fMRI method was a completely reliable means of truth verification and that Semrau’s results supported his lack of fraudulent intent.
- The United States moved in limine to exclude Laken’s testimony and the fMRI results under Federal Rule of Evidence 702 and the Daubert standard (and also raised Rule 403 concerns).
- Scientific literature submitted to the court indicated fMRI deception research remained largely experimental and not ready for courtroom use; in addition, the real-world error rate for fMRI-based lie detection was not established.
- A magistrate judge issued a Report and Recommendation addressing admissibility and recommending exclusion; the district judge adopted the recommendation, barring the fMRI lie-detection evidence at trial.

## Issues

1. Whether proposed fMRI-based lie-detection testimony and test results satisfy Federal Rule of Evidence 702 and Daubert’s reliability and relevance requirements when offered to show the defendant lacked intent to defraud.
2. Whether the fMRI lie-detection evidence should be excluded under Rule 403 because any probative value is substantially outweighed by the danger of unfair prejudice, confusion, or misleading the jury, including by intruding on the jury’s role in judging credibility.

## Decision

- The court granted the government’s motion to exclude Dr. Laken’s fMRI lie-detection testimony and the related fMRI test results.
- The court concluded the proffer failed Rule 702 and Daubert because the methodology was not shown to be sufficiently reliable for forensic use, including because the error rate was not known and published research did not establish courtroom-level validity.
- The court also determined the evidence posed substantial Rule 403 risks, including that jurors might assign excessive weight to brain-scan evidence and that the testimony functioned as “human lie detector” evidence directed at credibility.

## Legal Principles

- Under Federal Rule of Evidence 702, the proponent of scientific testimony must show the testimony is based on sufficient facts or data, is the product of reliable principles and methods, and reflects a reliable application of those methods to the case.
- Under Daubert, courts consider factors such as testability, peer review and publication, known or potential error rate, standards controlling the technique’s operation, and general acceptance in the relevant scientific community; the focus is on reliability and fit for the use proposed in litigation.
- Evidence drawn from controlled studies may be excluded where the research setting differs materially from the setting at issue in criminal adjudication (for example, low-stakes laboratory tasks versus high-stakes, motivated subjects), creating a mismatch between the science and the question the jury must decide.
- Where the error rate for a technique is unknown or not established for the use offered, that gap weighs strongly against admission under Rule 702 and Daubert.
- Even if evidence has some relevance, Rule 403 permits exclusion when scientific-sounding testimony or visual displays risk misleading the jury, confusing the issues, or creating unfair prejudice by giving an appearance of certainty beyond what the method can support.
- Courts generally do not permit experts to testify directly that a defendant or witness was truthful on a disputed matter, because credibility determinations are for the jury and such testimony can displace the jury’s role.

## Conclusion

In *United States v. Semrau*, the court excluded Cephos’s fMRI lie-detection evidence offered to show the defendant lacked intent to defraud, finding the proponent failed to demonstrate sufficient scientific reliability and case-specific fit under Rule 702 and *Daubert*—including because the technology’s forensic error rate was not established—and further ruling that the evidence posed a substantial risk of misleading jurors and invading the jury’s role under Rule 403.


United States v. Semrau, 2010 WL 6845092