A new method of detecting explosives using X-rays could also develop tumors

While the obvious software may be to search airports for bombs and other gadgets and harmful substances, the results, described in Nature Communications immediately, could also help detect cracks and rust in buildings, and finally, it could possibly be used to establish an early stage. tumors.

The team of researchers, from UCL in London, hid small portions of explosives, including Semtex and C4, inside electrical gadgets such as laptops, hair dryers and telephones laptops. Gadgets were placed inside luggage along with toothbrushes, chargers and other common items to neatly replicate a traveller’s bag.

While normal x-ray machines hit the objects with uniform x-ray discipline, the group scanned the luggage using a custom-built machine containing masks – sheets of steel pierced with holes, which separate the beams into an array of smaller beams. .

Scans inside a bag. High is typical, backside is microradian scattering.


Because the beams transmitted through the bag and its contents, they were scattered at angles as small as one microradian (about a 20,000th as large as a level). from a specific sample of angle settings.

AI is exceptionally good at retrieving these supplies even after they’re hidden in different objects, says lead designer Sandro Olivo, of UCL’s Medical Physics and Biomedical Engineering Division. “Even if we hide a small amount of explosive somewhere, because there may be a bit of texture in the middle of many different problems, the algorithm will find it.”

comparison between the conventional technique and the dispersion technique
Standard methodology (left) versus actual dissemination method.


The algorithm was able to appropriately establish explosives in each experiment conducted under examination conditions, although the team recognized that it would be unrealistic to rely on such a high degree of accuracy in larger research. that looked more like real-world situations.

The method is also used in medical functions, including screening for most cancers, the group believes. Although researchers only have to check whether the method could effectively differentiate the sensation of a tumor from surrounding healthy breast tissue, for example, they are excited about the possibility of detecting very small tumors that would not have been previously detected behind the rib of an affected person. cage.

“I would like to do it one day,” he says. “If we achieve an identical price of success in detecting texture in tumors, the potential for early analysis is significant.”

“This latest work from the UCL teams offered here looks extremely promising. It combines new X-ray imaging with AI and has primary potential for the extraordinarily difficult tasks of hazard detection in carry-on baggage and CND functions similar to crack detection,” says Kevin Wells, Affiliate Professor at the College of Surrey.

“Detecting most cancers involves its own set of personal challenges and we stay awake to see progress in this space sooner or later.”

Replace: The article has been updated with an extended citation.

Source link

Comments are closed.