All of these techniques measure the variations in some physiological responses of subjects in the course of answering various questions by a trained examiner, under the theory that these physiological variations are directly related to the subject's deception or truthfulness in answering the questions posed. Voice stress analysis theorizes that anxiety related to deception will be detectable by slight fluctuations in the subject's vocal recording. Studies of this technique have concluded that this theory is not valid. Thermal imaging measures changes in regional facial blood flow, particularly around the eyes. The goal of thermal imaging is to measure changes in blood flow related to the fright/flight response mediated by the sympathetic nervous system. Measurement is done through a camera that is sensitive to changes in temperature. Research on this method of detecting deception is still in its infancy, but is very promising, as it suggests that facial thermal measures may be effectively combined with polygraph measures in an examination. Pupillometry, the study of changes in pupil size and movement, is theorized to be relevant to detecting deception because pupil dilation can result from sympathetic nervous system stimulation or suppression of the parasympathetic nervous system, a phenomenon that has been relevant to lie-detection research. This technique has yet to establish significant reliability, because other factors besides deception can influence pupil size. Functional magnetic resonance imaging (MRI) involves the measurement or observation of increased blood flow to the active areas of the brain through the recording of a functional MRI image. The author reports that research in her lab has shown a relationship between functional MRI activity in the anterior cinbulate, middle, and superior frontal region of the brain and medial temporal gyri and deceptive responses. The author concludes that the polygraph is still the best measure of deception, but other techniques are also under scrutiny. She advises that these techniques are only as good as the theories that support them; i.e., there must be a reliable connection between what is being measured in the psychophysiological response of the subject and the subject's deception. 55 references