Raina Pang
Sep 16, 2011

Thermal imaging, lie detection and nervous flyers

Created at Sep 16, 2011, 3:15pm PDT

For humans, the ability to detect a lie accurately is advantageous in both professional and personal situations. Our fascination with accurate lie detection can be seen in pop culture with polygraphs being assessed by “MythBusters” and in crime dramas like “Lie to Me.” The principle behind lie detector tests is that deception is commonly associated with the physiological changes that underlie anxiety and nervousness. Traditionally, one thinks of the polygraph test as the main method of lie detection. Polygraphs measure physiological indices such as blood pressure, heart rate and skin conductance. Individuals with large physiological changes between control and relevant questions are considered deceitful. Because these measures require sensor attachment, the polygraph is impractical for large security scanning or for covert observations. However, assessing deceitfulness would be highly advantageous in these contexts.

It may be that lies can be detected by the face alone. Researchers at the University of Bradford are attempting to create an algorithm that can detect lying from changes in facial expressions and thermal regulation of the face. This technology is based on the notion that emotions are expressed through subtle changes of expressions and facial blood flow. Initial trials in an interview setting have determined this technique to be accurate in two out of three individuals. The creators are still modifying the technique, but a trial run will be implemented in UK airports later this year. Additionally, the creators believe their methods might prove useful in police interrogations or even job interviews.

While thermal imaging as lie detection would have a clear advantage over the polygraph test for large-scale security scanning, a number of issues need to be clarified. Under controlled settings, this technique is reported to have a 66 percent accuracy rating. Although researchers believe that high-stress situations, such as going through security checkpoints at airports, would increase the accuracy, real-world differences in personality and temperaments may create difficulties in accurate assessment. Many of the specific facial measures attributed to lying are also associated with nervousness in general; therefore, thermal imaging may show an increase in false positives related to individuals who are afraid of flying or mistrust airport personnel. 

The direction of error, a false positive or false negative, would have serious implications for the widespread use of thermal imaging lie detection in airports. A large number of false positives would inconvenience many flyers and could also potentially create a feedback loop in which increased anxiety would increase the number of false positives. A high false positive rate would also require a large amount of manpower to determine true threat and would invalidate the efficacy of this technology. While false positives are inconvenient, a high rate of false negatives would be disastrous. Most people would agree that a 66 percent success rate when assessing threat is not much better than chance alone.  Furthermore, the success rate has been measured only in control settings, so it remains unclear how effective these techniques would be when measuring sociopaths, pathological liars or countermeasures.

Generally, the scientific community remains skeptical of the methods used to detect lies. Yet, although many scientists reject polygraphs, lie detection methods find extensive use in post-conviction supervision, interrogations and daytime talk shows. While covert monitoring would make large-scale lie detection possible, many unanswered questions about its accuracy and real-world applicability remain.