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 ...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.展开更多
The main goal of this study was to introduce a novel three-dimensional procedure to objectively quantify both inner and outer condylar remodelling on preoperative multi-slice computed tomography （MSCT） and postopera...The main goal of this study was to introduce a novel three-dimensional procedure to objectively quantify both inner and outer condylar remodelling on preoperative multi-slice computed tomography （MSCT） and postoperative cone-beam computed tomography （CBCT） images. Second, the reliability and accuracy of this condylar volume quantification method was assessed. The mandibles of 20 patients （11 female and 9 male） who underwent bimaxillary surgery were semi-automatically extracted from MSCT/CBCT scans and rendered in 3D. The resulting condyles were spatially matched by using an anatomical landmark-based registration procedure. A standardized sphere was created around each condyle, and the condylar bone volume within this selected region of interest was automatically calculated. To investigate the reproducibility of the method, inter- and intra-observer reliability was calculated for assessments made by two experienced radiologists twice five months apart in a set of ten randomly selected patients. To test the accuracy of the bone segmentation, the inner and outer bone structures of one dry mandible, scanned according to the clinical set-up, were compared with the gold standard, micro-CT. Thirty-eight condyles showed a significant （P〈O.05） mean bone volume decrease of 26.4%_ 11.4% （502.9 mm3＋ 268.1 mm3）. No significant effects of side, sex or age were found. Good to excellent （ICC〉 0.6） intra- and inter-observer reliability was observed for both MSCT and CBCT. Moreover, the bone segmentation accuracy was less than one voxel （0.4 mm） for MSCT （0.3 mm __. 0.2 mm） and CBCT （0.4 mm _ 0.3 mm）, thus indicating the clinical potential of this method for objective follow-up in pathological condylar resorption.展开更多
在多层多匝矩形截面螺线管轴向分量磁场解析表达式难以求得的情况下,利用单匝矩形线圈及单层多匝矩形截面螺线管轴向分量磁场的解析表达式,采用切片求和的方法,求得了可用于软件编程的矩形截面螺线管轴向分量磁场分布的表达式,并编写了...在多层多匝矩形截面螺线管轴向分量磁场解析表达式难以求得的情况下,利用单匝矩形线圈及单层多匝矩形截面螺线管轴向分量磁场的解析表达式,采用切片求和的方法,求得了可用于软件编程的矩形截面螺线管轴向分量磁场分布的表达式,并编写了相应的Matlab计算程序.最终,利用所编写的计算程序,对各类小型加速器中应用较多的束流轨道校正磁铁轴向分量磁场的分布进行了计算,并与三维静态电磁场计算软件CST EM Studio的模拟结果进行了对比,两者符合较好.展开更多
基金supported by the National Natural Science Foundation of China(Nos.61771279,11435007)the National Key Research and Development Program of China(No.2016YFF0101304)
文摘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.
基金the Coordination for the Improvement of Higher Education Personnel(CAPES)programmeScience without borders from Brazilian governmentthe Research Foundation Flanders(FWO)from Flemish government for the fellowship support
文摘The main goal of this study was to introduce a novel three-dimensional procedure to objectively quantify both inner and outer condylar remodelling on preoperative multi-slice computed tomography （MSCT） and postoperative cone-beam computed tomography （CBCT） images. Second, the reliability and accuracy of this condylar volume quantification method was assessed. The mandibles of 20 patients （11 female and 9 male） who underwent bimaxillary surgery were semi-automatically extracted from MSCT/CBCT scans and rendered in 3D. The resulting condyles were spatially matched by using an anatomical landmark-based registration procedure. A standardized sphere was created around each condyle, and the condylar bone volume within this selected region of interest was automatically calculated. To investigate the reproducibility of the method, inter- and intra-observer reliability was calculated for assessments made by two experienced radiologists twice five months apart in a set of ten randomly selected patients. To test the accuracy of the bone segmentation, the inner and outer bone structures of one dry mandible, scanned according to the clinical set-up, were compared with the gold standard, micro-CT. Thirty-eight condyles showed a significant （P〈O.05） mean bone volume decrease of 26.4%_ 11.4% （502.9 mm3＋ 268.1 mm3）. No significant effects of side, sex or age were found. Good to excellent （ICC〉 0.6） intra- and inter-observer reliability was observed for both MSCT and CBCT. Moreover, the bone segmentation accuracy was less than one voxel （0.4 mm） for MSCT （0.3 mm __. 0.2 mm） and CBCT （0.4 mm _ 0.3 mm）, thus indicating the clinical potential of this method for objective follow-up in pathological condylar resorption.
文摘在多层多匝矩形截面螺线管轴向分量磁场解析表达式难以求得的情况下,利用单匝矩形线圈及单层多匝矩形截面螺线管轴向分量磁场的解析表达式,采用切片求和的方法,求得了可用于软件编程的矩形截面螺线管轴向分量磁场分布的表达式,并编写了相应的Matlab计算程序.最终,利用所编写的计算程序,对各类小型加速器中应用较多的束流轨道校正磁铁轴向分量磁场的分布进行了计算,并与三维静态电磁场计算软件CST EM Studio的模拟结果进行了对比,两者符合较好.
