How do you think the new GigE standards will influence the machine vision industry?
Respond or ask your question now!
By Lee J. Nelson
Research conducted by Dr. Daniel Langleben in the Department of Psychiatry, University of Pennsylvania Health System (Philadelphia) became the basis of a patent submission. Dr. Langleben—who was the first to show that lies and truths appear differently in brain scans—found increased activity in three areas of the conscious cortex and, from that, identified liars with 75 to 86 percent accuracy. The University licensed the then patent-pending rights to Joel Huizenga in exchange for an equity position in the company he formed (No Lie MRI, Inc., Tarzana, Calif.) to accomplish the technology-transfer. The U.S. Patent and Trademark Office (Alexandria, Va.) granted the patent in 2005. Today, No Lie MRI objectively and reliably measures intent, prior knowledge and deception with proprietary fMRI human brain mapping. The firm's VeraSource division develops the underlying analysis software while Veracity Sciences, another branch of No Lie MRI, focuses on implementation issues.
Despite the above achievements, practical obstacles impede widespread adoption of fMRI-based lie detection: the equipment is massive and, therefore, not transportable. Any head movement—let alone one's outright refusal to be scanned—negatively impacts image quality. In response, University of Pennsylvania Professor of Biochemistry and Molecular Biophysics, Britton Chance, designed an instrument to record much of the same brain activity as fMRI; but, it could fit in a briefcase and be deployed on an unwilling subject. To do so, Chance championed near-infrared light which passes harmlessly through the forehead and skull, just penetrating cortical tissue surfaces. There, the light bounces off the same variations in blood-flow as are tracked by fMRI. When it reemerges from the cranium, optical sensors capture the light, appropriately filter it and generate viewable images.
By monitoring blood oxygenation and volume in the prefrontal cortex, such functional near-infrared imaging (fNIR) returns different spectra (ranging from 700-1000 nm), indicating relative deoxy- and oxyhemoglobin concentrations as well as changes in the reduction/oxidation state of the metabolic enzyme, Cytochrome C oxidase.
Chance's first prototype consisted of a headband, studded with light-emitting diodes and silicon-diode sensors. Using it in a 2004 trial, Dr. Scott Bunce (Drexel University, Philadelphia) correctly detected lying among subjects 95 percent of the time; comparing favorably with fMRI. Since it can be worn as a manageable, minimally intrusive device, fNIR has the capacity to operate in real-life situations and everyday environments.
The next phase is to engineer a system for discreet remote sensing of the brain. Funded by the Office of Naval Research (Arlington, Va.), Chance and collaborators are working to replace the headpiece with an invisible laser and ultrasensitive photon collector. Their hope is to discern the neural signature of deception from across a room.