Slice-wise reconstruction for low-dose cone-beam CT using a deep residual convolutional neural network

Because of the growing concern over the radiation dose delivered to patients, X-ray cone-beam CT(CBCT) imaging of low dose is of great interest. It is difficult for traditional reconstruction methods such as Feldkamp to reduce noise and keep resolution at low doses. A typical method to solve this problem is using optimizationbased methods with careful modeling of physics and additional constraints. However, it is computationally expensive and very time-consuming to reach an optimal solution. Recently, some pioneering work applying deep neural networks had some success in characterizing and removing artifacts from a low-dose data set. In this study,we incorporate imaging physics for a cone-beam CT into a residual convolutional neural network and propose a new end-to-end deep learning-based method for slice-wise reconstruction. By transferring 3D projection to a 2D problem with a noise reduction property, we can not only obtain reconstructions of high image quality, but also lower the computational complexity. The proposed network is composed of three serially connected sub-networks: a cone-to-fan transformation sub-network, a 2D analytical inversion sub-network, and an image refinement sub-network. This provides a comprehensive solution for end-to-end reconstruction for CBCT. The advantages of our method are that the network can simplify a 3D reconstruction problem to a 2D slice-wise reconstruction problem and can complete reconstruction in an end-to-end manner with the system matrix integrated into the network design. Furthermore, reconstruction can be less computationally expensive and easily parallelizable compared with iterative reconstruction methods.

The applicability of vortex identification methods for complex vortex structures in axial turbine rotor passages

The complex vortex structures in the flow around turbine rotor passages, with weak or strong, large or small vortices, interacting with each other, often generate most of aerodynamic loss in turbomachines. Therefore, it is important to identify the vortex structures accurately for the flow field analysis and the aerodynamic performance optimization for turbomachines. In this paper, by using 4 vortex identification methods (the Q criterion, the Q method, the Liutex method and the Q -Liutex method), the vortices are identified in turbine rotor passages. In terms of the threshold selection, the results show that the D method and the Q -Liutex method are more robust, by which strong and weak vortices can be visualized simultaneously over a wide range of thresholds. As for the display consistency of the vortex identification methods and the streamlines, it is shown that the Liutex method gives results coinciding best with the streamlines in identifying strong vortices, while the Q -Liutex method gives results the most consistent with the streamlines in identifying weak vortices. As to the relationship among the loss, the vortices and the shear, except for the Q criterion, the other three methods can distinguish the vortical regions from the high shear regions. And the flow losses in turbine rotor passages are often related to high shear zones, while there is a small loss within the core of the vortex. In order to obtain the variation of vortices in the turbine rotor passages at different working points, the Liutex method is applied in 2 cases of a turbine with different angles of attack. The identification results show that the strengths of the tip leakage vortex and the upper passage vortex are weaker and the distance between them is closer at a negative angle of attack. This indicates that the Liutex method is an effective method, and can be used to analyze the vortex structures and their evolution in turbine rotor passages.