Jing Zou

Cone-beam computed laminography (CBCL) is widely used in non-destructive testing of plate-like samples due to flexible scanning geometry and fine image details. However, the Fourier domain defects lead to superimposed artifacts, which significantly degrade the image quality. This paper aims to assess the data distribution of CBCL projections in the Fourier domain, and to precisely delineate the Fourier defective volume. We derive the mapping scheme from CBCL projections to the Fourier domain using the Generalized Fourier Slice Theorem. Then, practical defect delineation methods are proposed for the cases of CBCL global scanning (i.e., non-truncated projections) and local scanning (i.e., truncated projections). Simulation experiments have been conducted to validate the proposed methods, indicating that the data incompleteness of CBCL is related not only to the scanning parameters but also to the shape and orientation of the object. Moreover, a de-artifacting algorithm based on Fourier priori information has been preliminarily implemented, which shows huge potential in efficiency and image quality over conventional SART-based algorithms.

BACKGROUND: Projection Domain Decomposition (PDD) is a dual energy reconstruction method which implements the decomposition process before image reconstruction. The advantage of PDD is that it can alleviate beam hardening artifacts and metal artifacts effectively as energy spectra estimation is considered in PDD. However, noise amplification occurs during the decomposition process, which significantly impacts the accuracy of effective atomic number and electron density. Therefore, effective noise reduction techniques are required in PDD. OBJECTIVE: This study aims to develop a new algorithm capable of minimizing noise while simultaneously preserving edges and fine details. METHODS: In this study, a denoising algorithm based on low rank and similarity-based regularization (LRSBR) is presented. This algorithm incorporates the low-rank characteristic of tensors into similarity-based regularization (SBR) framework. This method effectively addresses the issue of instability in edge pixels within the SBR algorithm and enhances the structural consistency of dual-energy images. RESULTS: A series of simulation and practical experiments were conducted on a dual-layer dual-energy CT system. Experiments demonstrate that the proposed method outperforms exiting noise removal methods in Peak Signal-to-noise Ratio (PSNR), Root Mean Square Error (RMSE), and Structural Similarity (SSIM). Meanwhile, there has been a notable enhancement in the visual quality of CT images. CONCLUSIONS: The proposed algorithm has a significantly improved noise reduction compared to other competing approach in dual-energy CT. Meanwhile, the LRSBR method exhibits outstanding performance in preserving edges and fine structures, making it practical for PDD applications.

To address the limited computing resources and real-time data processing requirements of mobile devices in the mobile edge network environment, this paper proposes a deep reinforcement learning-based edge-cloud collaborative computing offloading method. Specifically, considering computational resources, bandwidth, and task offloading strategies comprehensively, we formulate the problem of minimizing task execution delay, energy consumption, and cost for all mobile devices. Based on this, we design an online coordinated offloading algorithm using deep reinforcement learning, which transforms the cost minimization problem into an optimal offloading strategy problem under Markov decision process framework. The best task offloading strategy is obtained by utilizing experience replay for solution. Experimental results demonstrate that the proposed algorithm exhibits good convergence and outperforms other methods in terms of system performance.

Additive manufacturing (AM) is a layer-by-layer manufacturing method which is growing considerably in recent years. Inner dimensional measurement based on Computed Tomography (CT) image is an irreplaceable method for quality control in AM. However, for special AM components, limited-angle CT imaging is the only method because of physical limitations. CT reconstruction image with ghost artifacts through incomplete data can’t be used to implement dimensional measurement. To deal with this problem, a method based on encoder-decoder neural network is proposed and applied to characterize the dimensional parameter directly from CT sinogram without reconstruction. Experiment results showed that when characterizing AM component with length of side from 10 to 15mm, the relative accuracy is less than 18.26% and the relative accuracy is less than 10% in over 83.5% sample, which verifies the effectiveness of the method that can be applied in dimensional measurement.

Nowadays, with the rapid development of computed tomography (CT), the theoretical expansion of CT reconstruction has become the crucial ingredient to break through technical bottlenecks. This article supplements relevant knowledge of the method for constructing CT projection-wise filters. From the perspective of the mathematical principle of CT reconstruction, the second-order divided-difference back projection (SDBP) technique is firstly proposed, which is an implementation of accurate and efficient inverse Radon transform. On the basis of the SDBP technique, a brand new computational model for filter expressions is derived. The model reveals the correlation between the convolution kernel for data restoration and the filter expression and also specifies the principle of filter construction. On the proposed filter construction model, the decomposition form of filters is discovered for the first time. A series of basic filters are acquired, which can be used to compose filters in actual need. According to the superposition principle, the properties of a filter depend on the decomposed basic filters. The selection of the kernel function on the properties of basic filters clarifies the general rule of filter construction. These theories proposed in this article are of reference value for the filter optimization research in the CT reconstruction field.

Dual-energy computed tomography (CT) provides a better material classification capability than traditional CT. A scintillator is an essential component that determines the optical resolution, detective quantum efficiency, and energy separation of the dual-layer dual-energy CT system. In this paper, we propose a new dual-layer scintillator combination composed of CsI:Tl and LYSO:Ce for dual-energy CT imaging. The x-ray attenuation physical process of CsI:Tl and LYSO:Ce combination is simulated on Geant4. The energy distribution and luminescence of different thicknesses of the scintillator combination are investigated by simulation experiments. Then, the optimal thicknesses of the combination are designed based on simulation results. Furthermore, a dual-layer dual-energy CT system using the scintillator combination is built, and practical dual-energy imaging experiments are implemented in the system. Simulation results indicate that the CsI:Tl and LYSO:Ce combination can effectively realize dual-energy detection. The optimal thickness is 100 μm for CsI:Tl and 500 μm for LYSO:Ce. Practical experiments show that the dual-energy CT system using the CsI:Tl and LYSO:Ce combination has low noise. Materials of the metal mixture sample and the igneous rock sample are correctly separated using the dual-energy data obtained by the CsI:Tl and LYSO:Ce combination. Therefore, the CsI:Tl and LYSO:Ce combination is expected to be used in dual-energy CT imaging studies of complex materials, such as metals, bones, and rocks.

Linear Computed Laminography (LCL) is used to yield slice images of plate-like objects (PLO) for the advantage of short exposure time, high control precision and low cost. Shift and Add (SAA) is a widely used reconstruction algorithm for LCL. One limitation of SAA is that the reconstructed image of the in-focus layer (IFL) contains information from off-focus layers (OFL), resulting in inter-slice aliasing and blurring. In this paper, an Iterative Difference Deblurring (IDD) algorithm based on LCL is proposed to reduce the blur in reconstructed images. The core idea of the IDD algorithm is: contributions from OFL are subtracted from the projection data to remove the blur from the IFL. The corrected projections are then reconstructed using the SAA to remove the superimposed contributions of OFL from the IFL. An iterative approach is utilized to adjust a weighting factor applied during the subtraction stage. The results demonstrate that IDD algorithm can achieve PLO reconstruction in the LCL system under extremely sparse sampling conditions, and can effectively reduce the inter-slice aliasing and blurring.

Different scintillators have different x-rays attenuation characteristics. Many researchers are committed to developing scintillators with strong x-rays stopping power for industrial x-rays computed tomography (CT) imaging. But scintillators with strong x-rays stopping power may not always be the best choice. In this paper, we study the scintillator selection for improving material separation ability of low attenuation samples in x-rays imaging. Digital radiography (DR) experiments using cerium-doped lutetium yttrium oxyorthosilicate (LYSO: Ce, strong x-rays stopping power), cerium-doped gadolinium aluminum gallium garnet (GAGG: Ce, medium x-rays stopping power) and aluminum-doped zinc selenide (ZnSe: Al, weak x-rays stopping power) scintillators are implemented based on an optical coupling detection system. Material separation ability using DR images at 40-130 kVp x-rays sources are compared. Besides, CT imaging experiments of a Chinese herbal medicine using the three scintillators are completed. Material separation results of images obtained using ZnSe: Al is better than that using GAGG: Ce and LYSO: Ce. Thus ZnSe: Al is the best choice for material separation in x-rays CT imaging of low attenuation samples among the three scintillators.