Researchers at The University of Texas at Austin in the Mechanical Engineering's Nuclear and Radiation Program were recently published in the journal Nature for their work developing a method to detect small amounts of radioactive materials in layered objects, such as airport luggage.
Ph.D. student Andrew Gilbert, first author on paper profiled in the journal Nature.
Researchers at The University of Texas at Austin in the Mechanical Engineering's Nuclear and Radiation Program were recently published in the journal Nature for their work developing a method to detect small amounts of radioactive materials in layered objects, such as airport luggage. The profile entitled "X-ray method could improve nuclear screening" was published on April 18, 2014, after their article appeared in the Journal of Applied Physics.
University of Texas at Austin's Mechanical Engineering Ph.D. student Andrew Gilbert, first author and his advisor Assistant Professor Mark Deinert, in the Nuclear Engineering Program, along with researchers at the Department of Energy's Pacific Northwest National Laboratory (PNNL) in Richland, Washington have been able to visually detect radioactive materials using spectral x-ray detectors. Currently, detection of radioactive material isn't possible with standard airport scanning devices. The team coupled commercially available spectral X-ray detectors with a specialized algorithm than can improve detection of uranium and plutonium in small, layered objects such as luggage.
National Security
According to the International Atomic Energy Agency (IAEA) the greatest danger to nuclear security comes from terrorists acquiring sufficient quantities of plutonium or highly enriched uranium (HEU) to construct a crude nuclear explosive device. The IAEA also notes that most cases of illicit nuclear trafficking have involved gram-level quantities, which can be challenging to detect with most inspection methods.
Spectral detection compared to traditional X-ray
Spectral detectors were first developed for medical imaging to better differentiate between bone and soft tissue, and produce a color image, not a black image seen in traditional x-ray imaging. The spectrally sensitive detectors are akin to color vision, and this extra information can be used to make sensitive measurements of material composition. Different materials will shift the spectrum in slightly different ways, and the algorithm allows researchers to distinguish different radioactive materials, which are denser than other materials.
An Inverse Algorithm
"We first had to develop a computational model for how X-rays move through materials and how they are detected so that we could predict what an image will look like once the radiation passed through an object," said Andrew Gilbert a graduate student in Professor Deinert's group. "With that in hand, we applied an 'inverse algorithm,' varying the composition of the object until the predicted image matched the measured one. We also gave our algorithm additional details about density and other factors—a process called 'regularization'—to adaptively enhance its ability to discriminate materials." It can be difficult to obtain accurate results if the radiographic signal contains noise. Regularization improves the optimization results, and amounts to adding a constraint or additional information to the optimization problem, which evens the solution by penalizing for sharp variations between neighboring pixels.
Preliminary testing of layered materials
They demonstrated the approach using single-view synthetic radiographs of objects made from layers of plutonium, cotton, steel, lead, aluminum, and copper. Image noise and object scattered x-rays are the dominant factors affecting the accuracy of the estimates.
Now that the inverse algorithm method has been shown to help X-rays detect nuclear materials in luggage and other small objects, Deinert said that his team will next expand the concept to improve detection on a larger scale. "We plan to apply the algorithm to high-energy X-ray systems that could be used for verification of arms-reduction treaties," he said.
The article in the Journal of Applied Physics
The original paper titled, "Non-invasive material discrimination using spectral X-ray radiography" authored by A.J. Gilbert, B.S. McDonald, S.M. Robinson, K.D. Jarman, T.A. White and M.R. Deinert was published in the Journal of Applied Physics on April 15, 2014. Published by the American Institute of Physics, it is an influential international journal that publishes significant new experimental and theoretical results of applied physics research.
- Journal of Applied Physics 115 , 154901 (2014); doi: 10.1063/1.4870043
- View online
- View Table of Contents
- Published by the AIP Publishing
