研究課題/領域番号 
18F18763

研究種目 
特別研究員奨励費

配分区分  補助金 
応募区分  外国 
研究分野 
量子ビーム科学

研究機関  国立研究開発法人理化学研究所 
受入研究者 
大竹 淑恵 国立研究開発法人理化学研究所, 光量子工学研究センター, チームリーダー (50216777)

外国人特別研究員 
YAN MINGFEI 国立研究開発法人理化学研究所, 光量子工学研究センター, 外国人特別研究員

研究期間 (年度) 
20181109 – 20210331

研究課題ステータス 
交付 (2019年度)

配分額 *注記 
2,100千円 (直接経費: 2,100千円)
2020年度: 700千円 (直接経費: 700千円)
2019年度: 900千円 (直接経費: 900千円)
2018年度: 500千円 (直接経費: 500千円)

キーワード  Nondestructive test / Neutron / Linear scanning CT / Image reconstruction / Algorithm 
研究実績の概要 
We have established the simulation model for rotation+linear scanning CT in experiment condition based on the neutron beam condition of Riken acceleratordriven compact neutron source (RANS) with Geant4 code, and established system matrix calculation method with line integral model. Then we applied the sparsity reconstruction algorithm to rotation+linear scanning CT reconstruction with the simulated projection data and calculated system matrix. The results show 1cm space resolution image can be obtained by rotation and linear scanning method with pixeltype fast neutron detector (pixel size: 2cm). The experiment research on RANS is under preparation. Furthermore, we have implemented the linear scanning CT reconstruction with the projection date from simulation with Geant4 code, and current research results indicate satisfactory reconstructed image can be obtained by linear scanning method. In addition, we have started to develop the new reconstruction algorithm incorporating the finite spot size of neutron source into the reconstruction process. According to our current progress, we have determined to attend the 8th international meeting of the union for compact acceleratordriven neutron sources(UCANS8) holding in Paris in July to show our recent research results about rotation +linear scanning CT. After we finished the rotation +linear scanning CT experiment on compact neutron source, we will publish a journal paper about the simulation, experiment and reconstruction for rotation+linear scanning CT.

現在までの達成度 (区分) 
現在までの達成度 (区分)
2: おおむね順調に進展している
理由
1. We have adjusted the research contents a little bit according to the current experiment condition. In the contents, we have added the rotation +translation scanning CT as the transition between rotation CT (like medical CT) and linear scanning CT. Currently, we have established the simulation model for such kind of CT by Geant4 code according to the experiment condition, calculated the corresponding system matrix, and realized the reconstruction with the recent sparsity CT image reconstruction algorithm. In addition, we have start to test the pixeltype fast neutron detector for the following experiment, and the experiment sample preparation is also under consideration. After we finished the experiment, we plan to publish one journal paper about simulation, experiment and reconstruction for roatation+linear scanning CT. 2. We have conducted the primary simulation and reconstruction for linear scanning CT. We have established the simulation model for such kind of CT by Geant4 code with ideal condition, calculated the corresponding system matrix, and realized the reconstruction with recent sparsity reconstruction algorithm. The extensive research on such kind of CT is undergoing, in which we are developing the new CT image reconstruction algorithm for linear scanning CT when considering the finite spot size source. After we realized the algorithm mentioned above, we plan to conduct experiment research on linear scanning CT. We plan to publish a journal paper about simulation, experiment and reconstruction algorithm for linear scanning CT.

今後の研究の推進方策 
1. The experiment of rotation +linear scanning CT for concrete sample will be implemented on compact neutron source RANS according to the plan. 2. The extensive research on linear scanning CT regarding of the influence of scanning parameters(such as scanning direction, translation step and so on) on accuracy of reconstructed image will be conducted. 3. Many reconstruction algorithms are based on the mathematics model in which the source is regarded as ideal point source for cone/fan beam without considering its size, and the projection data is not overlapped. However, since the finite spot size of neutron source has serious influence on the space resolution of reconstructed image in linear scanning CT, it should be incorporated into the reconstruction algorithm. When the finite spot size is considered, the projection data is overlapped in CT image reconstruction process, and we need to construct the new the mathematics model as well as corresponding new reconstruction algorithm. This is the main problem needing to be solved in the following research.
