A novel detection system inspired by the classic game ‘Tetris’ has the potential to create affordable and precise radiation detectors for surveillance at nuclear facilities. The aftermath of the Fukushima Daiichi Nuclear Power Plant disaster in Japan in 2011 and the continuous risk of a potential radiation release from the Zaporizhzhia nuclear complex in the Ukrainian conflict zone have highlighted the importance of finding effective and dependable methods for detecting and monitoring radioactive isotopes. Additionally, routine activities at nuclear reactors, as well as the mining and conversion of uranium into fuel rods, and the management of spent nuclear fuel, also call for reliable detection and monitoring techniques.o require monitoring of radioisotope release.
Researchers from MIT and the Lawrence Berkeley National Laboratory (LBNL) have developed a computational method for creating simple, efficient sensor configurations that can accurately identify the direction of a distributed radiation source. They also showed that by moving the sensor to obtain multiple readings, they can determine the specific location of the source. The idea for their innovative approach was inspired by the well-known computer game “Tetris.”
The team’s discoveries, which have the potential to be applied to detector technology in general, were a result of their research.The detection of radiation, including gamma rays, is typically accomplished using semiconductor materials like cadmium zinc telluride. These materials generate an electrical response when they are hit by high-energy radiation. However, because radiation can easily pass through objects, it is challenging to determine the direction from which the signal originated. This issue, along with solutions for other types of radiation, is explored in a paper published in Nature Communications by MIT professors Mingda Li, Lin-Wen Hu, Benoit Forget, and Gordon Kohse; graduate students Ryotaro Okabe and Shangjie Xue; research scientist Jayson Vavrek SM ’16, PhD ’19 at LBNL; and various others at MIT and Lawrence Berkeley.The article discusses the challenges of detecting radiation sources using simple counting devices like Geiger counters. These devices only produce a click sound when they detect radiation, without providing detailed information about the energy or type of radiation. As a result, finding the source of radiation requires physically moving around to locate the maximum sound, similar to how handheld metal detectors work. However, this process can be risky as it requires the user to move closer to the source of radiation.
To address this challenge, researchers have developed a method to provide directional information from a stationary device without the need to get too close to the radiation source. This method involves using an array of detector grids along with a mask, which imprints a pattern on the array that varies depending on the direction of the source. An algorithm is then used to analyze the patterns and determine the direction of the radiation source.The interpretation of the various timings and strengths of signals received by each individual detector or pixel often results in a complicated design of detectors.
Typical detector arrays used to sense the direction of radiation sources are typically large and costly, consisting of at least 100 pixels arranged in a 10 by 10 array. However, the researchers discovered that using as few as four pixels arranged in the tetromino shapes from the “Tetris” game can achieve similar accuracy to the large, expensive systems. The key lies in accurate computerized reconstruction of the angles of arrival of the rays, based on the timing of the signal detection by each sensor and th rnrnThe AI-guided study of simulated systems reconstructed the relative intensity detected by each pixel. The researchers experimented with different configurations of four pixels, including square, S-shaped, J-shaped, and T-shaped. Through repeated experiments, they found that the most precise results were provided by the S-shaped array. This array accurately provided directional readings within about 1 degree, outperforming the square and the other irregular shapes. The inspiration for this approach came from the game ‘Tetris.’ Placing an insulating material such as a lead sheet between the pixels is crucial for making the system work.The purpose of using a small detector is to enhance the contrast between radiation readings received from different directions. The spacing between the pixels in these simplified arrays performs the same role as the more complex shadow masks used in larger-array systems. According to Okabe, the lead author of the study, less symmetrical arrangements offer more valuable information from a small array. “The advantage of using a small detector lies in terms of engineering costs,” he explains. The individual detector elements are costly, usually made of cadmium-zinc-telluride, or CZT, and all of theThe connections that carry information from the pixels become more complex as the detector gets smaller and simpler, making it better for various applications, according to Li. Many simplified arrays for radiation detection are only effective when the radiation comes from a single localized source, but the “Tetris”-based version can handle multiple or spread-out sources well, as noted by Xue, who co-authored the work. In a single-blind field test at the Berkeley Lab with a real cesium radiation source, led by Vavrek,Researchers at MIT conducted a test where they were able to accurately determine the direction and distance to a source, even when the ground-truth source location was unknown.
According to co-author Forget, a professor of nuclear engineering at MIT, radiation mapping is crucial for the nuclear industry as it helps in quickly locating radiation sources and ensuring the safety of everyone involved.
Co-lead-author Vavrek mentioned that their study focused on gamma-ray sources and they developed computational tools to extract directional information.A small number of pixels are “much, much more general.” It is not limited to specific wavelengths, and can also be used for neutrons, or even other types of light, such as ultraviolet light, according to Hu, a senior scientist at MIT Nuclear Reactor Lab.
Other members of the research team include Ryan Pavlovsky, Victor Negut, Brian Quiter, and Joshua Cates at Lawrence Berkely National Laboratory, and Jiankai Yu, Tongtong Liu, and Stephanie Jegelka at MIT. The study received support from the U.S. Department of Energy.
Journal Reference:
- Ryotaro Okabe, Shangjie Xue, Jayson R. Vavrek, Jiankai Yu, Ry rnPavlovsky, Victor Negut, Brian J. Quiter, Joshua W. Cates, Tongtong Liu, Benoit Forget, Stefanie Jegelka, Gordon Kohse, Lin-wen Hu, Mingda Li. Tetris-inspired detector with neural network for radiation mapping. Nature Communications, 2024; 15 (1) DOI: 10.1038/s41467-024-47338-w